{
  "schema_version": "wide_deep_progress_snapshot_v7",
  "status": {
    "label": "21轮训练完成；WT、DT待补测，WDT测试进行中",
    "page_is_a_live_monitor": false,
    "observed_at_utc": "2026-11-02T09:29+00:00",
    "observed_at_beijing": "2026-11-02 17:29",
    "original_formal_training_started": true,
    "original_active_revision": "teacher_cluster_deep_history_v6",
    "active_arm": "WDT",
    "active_round": null,
    "active_phase": "common_test",
    "active_answers_saved": 44,
    "active_answers_required": 90,
    "completed_learning_rounds": 12,
    "original_required_learning_rounds": 12,
    "completed_original_full_training_arms": [
      "WT",
      "DT",
      "WDT",
      "R"
    ],
    "completed_experiment_rounds": 3,
    "extension_release_ready": true,
    "extension_deployed": true,
    "extension_training_started": true,
    "new_method_results_available": false,
    "final_evaluation_started": true,
    "final_evaluation_completed": false,
    "billing_stopped_confirmed": false,
    "reported_monitor_error": null,
    "extension_prepared_and_attached": true,
    "extension_controller_started": true,
    "extension_waits_for_original_round3_checkpoints": false,
    "active_answer_count_available_in_snapshot": true,
    "active_revision": "curriculum_repair_release_v7",
    "active_phase_label": "WDT共同测试",
    "active_answers_origin": "main_common_test_WDT",
    "active_new_main_answers_saved": 0,
    "active_imported_main_learning": false,
    "completed_learning_rounds_scope": "original_four_groups",
    "original_training_complete": true,
    "extension_started_at_utc": "2026-11-01T22:05:55.929321+00:00",
    "extension_started_at_beijing": "2026-11-02 06:05:55",
    "training_started_definition": "本组训练流程已经开始；复用首轮记录不计为新增学习。",
    "extension_new_learning_started": true,
    "new_learning_updates_completed": 10,
    "main_test_answers_saved": 302,
    "auxiliary_test_answers_saved": 0,
    "active_answers_upper_bound": 90,
    "active_question_cap": 30,
    "extension_learning_results_available": true,
    "new_method_results_scope": "independent_final_test_comparison_not_yet_available",
    "latest_completed_repair": {
      "schema_version": "wide_deep_REPAIR_round3_summary_v1",
      "arm": "WDT_REPAIR",
      "round": 3,
      "status": "COMPLETE",
      "observed_at_utc": "2026-11-02T06:28:20.285682+00:00",
      "observed_at_beijing": "2026-11-02 14:28:20",
      "started_at_utc": "2026-11-02T04:46:39.165028+00:00",
      "completed_at_utc": "2026-11-02T06:22:35.663424+00:00",
      "completed_at_beijing": "2026-11-02 14:22:35",
      "questions": 44,
      "answers": 352,
      "seed_questions": 20,
      "seed_answers": 160,
      "seed_successes": 118,
      "previous_round_seed_successes": 125,
      "common_old_questions": 32,
      "common_old_answers": 256,
      "common_old_successes_previous": 159,
      "common_old_successes_current": 164,
      "review_questions": 12,
      "new_questions": 12,
      "effective_questions": 18,
      "all_correct_questions": 17,
      "all_wrong_questions": 9,
      "effective_new_questions": 4,
      "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
      "new_difficulty_by_kind": {
        "deep_down": {
          "all_failed": 0,
          "all_succeeded": 1,
          "effective_questions": 1,
          "minimum_actions_delta_range": [
            -5,
            -4
          ],
          "questions": 2,
          "required_chain_depth_delta_range": [
            -1,
            -1
          ]
        },
        "deep_up": {
          "all_failed": 1,
          "all_succeeded": 0,
          "effective_questions": 3,
          "minimum_actions_delta_range": [
            2,
            4
          ],
          "questions": 4,
          "required_chain_depth_delta_range": [
            1,
            1
          ]
        },
        "wide": {
          "all_failed": 2,
          "all_succeeded": 4,
          "effective_questions": 0,
          "minimum_actions_delta_range": [
            -1,
            1
          ],
          "questions": 6,
          "required_chain_depth_delta_range": [
            -1,
            0
          ]
        }
      },
      "signal_coverage": {
        "failed_questions": 20,
        "first_answer_failed_questions": 10,
        "independent_student_answers": 0,
        "later_failures_considered": 10,
        "mixed_source_questions": 14,
        "regressed_question_ids": [
          "generated:local_textcraft_188",
          "generated:local_textcraft_248",
          "generated:local_textcraft_280",
          "generated:local_textcraft_462",
          "generated:local_textcraft_502",
          "generated:local_textcraft_517"
        ],
        "source_questions": 32,
        "upward_sources_skipped_without_teacher_proposals": 1
      },
      "main_learning_updates": 1,
      "repair_learning_updates": 1,
      "new_learning_updates_completed": 2,
      "main_learning_complete": true,
      "repair_learning_complete": true,
      "main_model_saved": true,
      "repair_model_saved": true,
      "model_saved": true,
      "main_learning_reused": false,
      "new_branch_answers": 352,
      "collection_policy_id": "64bba71af88c1427cc676e23f26a83f3f6f70ead23698dae58c93bd283d64fe7",
      "main_policy_id": "f445a3e252ec14df358c1fc37bb26fd7c0756cce8034ad607926b0642b7330a7",
      "final_policy_id": "aeca99455724c3b1855ef6b0ed30ea2fce7dd20dfd01f7a907d386b0c2667ce3",
      "previous_round_final_model_matches_collection_model": true,
      "answers_collected_before_third_learning": true,
      "completed_main_learning_rounds_before_collection": 2,
      "answer_meaning": "第三轮答卷在本轮主练和纠错前收集，反映第二轮最终模型。第三轮最终效果待共同30题测试。",
      "effect_interpretation": "固定20道种子由125/160降至118/160；共同32道旧题由159/256升至164/256，两个范围的变化方向不同。",
      "repair": {
        "accepted_by_mode": {
          "prevention": 31,
          "recovery": 15
        },
        "accepted_proposals": 46,
        "optimizer_steps": 1,
        "proposal_count": 69,
        "rejection_reasons": {
          "NO_SUCCESSFUL_RELEVANT_CORRECTION": 2,
          "SUFFIX_DID_NOT_COMPLETE_ORIGINAL_TASK": 21
        },
        "repaired_target_ids": [
          "generated:cluster_wdt_r1_s0",
          "generated:cluster_wdt_r1_s18",
          "generated:cluster_wdt_repair_r2_s12",
          "generated:cluster_wdt_repair_r2_s2",
          "generated:cluster_wdt_repair_r2_s3",
          "generated:cluster_wdt_repair_r2_s5",
          "generated:cluster_wdt_repair_r3_s21",
          "generated:cluster_wdt_repair_r3_s23",
          "generated:cluster_wdt_repair_r3_s5",
          "generated:cluster_wdt_repair_r3_s7",
          "generated:local_textcraft_188",
          "generated:local_textcraft_248",
          "generated:local_textcraft_280",
          "generated:local_textcraft_448",
          "generated:local_textcraft_462",
          "generated:local_textcraft_489",
          "generated:local_textcraft_502",
          "generated:local_textcraft_517"
        ],
        "selected_actions": 73,
        "original_target_count": 23,
        "original_goal_completed_proposals": 47,
        "accepted_target_count": 18,
        "completed_original_goal_without_accepted_actions": 1,
        "selected_correction_tokens": 1055,
        "original_target_cases": [
          {
            "task_id": "generated:local_textcraft_477",
            "label": "书架",
            "proposals": 3,
            "original_goal_completed_proposals": 0,
            "accepted_proposals": 0
          },
          {
            "task_id": "generated:local_textcraft_489",
            "label": "完整红旗",
            "proposals": 3,
            "original_goal_completed_proposals": 3,
            "accepted_proposals": 3
          },
          {
            "task_id": "generated:local_textcraft_188",
            "label": "安山岩",
            "proposals": 3,
            "original_goal_completed_proposals": 3,
            "accepted_proposals": 2
          }
        ],
        "original_goal_completion_definition": "原记录 judge_status 为 PASS；与是否存在合格学习动作分别统计。",
        "accepted_proposals_definition": "完成原目标并通过改正动作筛选的方案。",
        "selected_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
        "student_independent_success_after_learning_verified": false
      },
      "final_common30_test_completed": false,
      "complete_round_seconds": 5756.498396,
      "stage_seconds": {
        "plan-round": 2566.931565,
        "train-round": 3189.53828
      },
      "pure_parameter_training_seconds": null,
      "timing_scope": "完整一轮，含出题、作答、主练、纠错、保存和阶段切换。",
      "stage_observation_intervals": {
        "remaining_answers_and_preparation_seconds_interval": [
          696.882401,
          743.968748
        ],
        "main_learning_plus_saving_seconds_interval": [
          450.458368,
          543.847943
        ],
        "repair_material_generation_and_validation_seconds_interval": [
          1892.11893,
          1977.927737
        ],
        "repair_learning_saving_and_exit_seconds_interval": [
          17.183427,
          56.689006
        ]
      },
      "stage_observation_interval_scope": "相邻观察确定的时间范围，不是纯模型计算时间；区间上限不能直接相加。",
      "question_results": [
        {
          "task_id": "generated:local_textcraft_188",
          "section": "seed",
          "kind": "root",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_215",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_248",
          "section": "seed",
          "kind": "root",
          "successes": 3,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_273",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_280",
          "section": "seed",
          "kind": "root",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_42",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_448",
          "section": "seed",
          "kind": "root",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_462",
          "section": "seed",
          "kind": "root",
          "successes": 3,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_477",
          "section": "seed",
          "kind": "root",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_480",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_489",
          "section": "seed",
          "kind": "root",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_502",
          "section": "seed",
          "kind": "root",
          "successes": 4,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_517",
          "section": "seed",
          "kind": "root",
          "successes": 6,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_328",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_357",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_100",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_72",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_45",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_190",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_355",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_r1_s18",
          "section": "review",
          "kind": "wide",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s4",
          "section": "review",
          "kind": "wide",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_r1_s5",
          "section": "review",
          "kind": "deep_down",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s10",
          "section": "review",
          "kind": "deep_down",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s5",
          "section": "review",
          "kind": "wide",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_r1_s1",
          "section": "review",
          "kind": "deep_down",
          "successes": 4,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s17",
          "section": "review",
          "kind": "deep_down",
          "successes": 6,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s3",
          "section": "review",
          "kind": "wide",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s0",
          "section": "review",
          "kind": "wide",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s12",
          "section": "review",
          "kind": "deep_up",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s14",
          "section": "review",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r2_s2",
          "section": "review",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s0",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s1",
          "section": "new",
          "kind": "deep_down",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s16",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s17",
          "section": "new",
          "kind": "deep_down",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s4",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s5",
          "section": "new",
          "kind": "deep_up",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s20",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s21",
          "section": "new",
          "kind": "deep_up",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s6",
          "section": "new",
          "kind": "wide",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s7",
          "section": "new",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s22",
          "section": "new",
          "kind": "wide",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_repair_r3_s23",
          "section": "new",
          "kind": "deep_up",
          "successes": 3,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        }
      ],
      "source_records": {
        "CAPTURE.json": {
          "bytes": 58292,
          "sha256": "481c8837533874f3fcc9d252a3779276d4967bba13aaf908273090373464d3c0"
        },
        "complete.json": {
          "bytes": 9988,
          "sha256": "94889594a144983fd0a0bd53ce66a79f38c72aeb7b8218d4be6868492535af1e"
        },
        "collection.json": {
          "bytes": 12803,
          "sha256": "05ccda0149e1ca38cabe05bb2db12793d5cda60a5853ef2ecaa99a684e720b6f"
        },
        "plan.json": {
          "bytes": 101855,
          "sha256": "cf4a8aac703816624a1ad4bc09cc51656228a6ff920b5947ab9f1286b220e0c8"
        },
        "main.json": {
          "bytes": 8330,
          "sha256": "80ea9fbc1874462bb70cd08e5b134e9ead463148db9d672c67b66e6020b6a759"
        },
        "repair.json": {
          "bytes": 952,
          "sha256": "a9aac9f467d3dd314e804ea9a0205150d5607d8ca7928fd4b9ec097e4a6008c5"
        },
        "repair_summary.json": {
          "bytes": 116040,
          "sha256": "3c50f288bb48e2ec1440abfa4794cfc601f79c6ba574a786780a675d4dbab041"
        },
        "main_model/COMPLETE.json": {
          "bytes": 337,
          "sha256": "9af78ea8fced779502f596ff16b3076f12dc05de0324b1110ac2a350e4649d99"
        },
        "repair_model/COMPLETE.json": {
          "bytes": 337,
          "sha256": "938db3362a51c2379bf21664056a7281fe2cfe67a92f2a47b8dad82bd397ea49"
        },
        "timing/WDT_REPAIR_plan-round_3_attempt_01.started.json": {
          "bytes": 1141,
          "sha256": "b9c492335e3fbf15cb8a3a29143d5b1d0416666fd571fce252f7f02e72208b5c"
        },
        "timing/WDT_REPAIR_plan-round_3_attempt_01.exit.json": {
          "bytes": 180,
          "sha256": "5dc4349ad5667c350da7433bf1d232e41b7b24498a636f6de6f52b4c60d5d19f"
        },
        "timing/WDT_REPAIR_train-round_3_attempt_01.started.json": {
          "bytes": 1148,
          "sha256": "76bf2a00a9f18084a8d6b25eb69acf9225e24c4cf8e08aaaffe85c9dc412edc5"
        },
        "timing/WDT_REPAIR_train-round_3_attempt_01.exit.json": {
          "bytes": 181,
          "sha256": "e6fe2939a08c4705ad4a143a9e4599d52f6d6b8a92c86fae887284913c6153e1"
        }
      },
      "parent_analysis": {
        "bytes": 8575,
        "sha256": "559e0778b28fff5ece438ae2b57f577dfd0d50c06e2ffe4829c427fd68a9e80b"
      }
    },
    "latest_completed_scale50": {
      "arm": "WDT_SCALE50",
      "round": 3,
      "status": "COMPLETE",
      "seed_questions": 50,
      "review_questions": 25,
      "new_questions": 9,
      "questions": 84,
      "answers": 672,
      "answers_per_question": 8,
      "effective_questions": 40,
      "effective_new_questions": 4,
      "all_succeeded_questions": 33,
      "all_failed_questions": 11,
      "new_main_learning_updates": 1,
      "final_model_saved": true,
      "repair_enabled": false,
      "new_question_breakdown": {
        "deep_down": {
          "all_failed": 0,
          "all_succeeded": 1,
          "effective_questions": 0,
          "minimum_actions_delta_range": [
            -10,
            -10
          ],
          "questions": 1,
          "required_chain_depth_delta_range": [
            -1,
            -1
          ]
        },
        "deep_up": {
          "all_failed": 0,
          "all_succeeded": 0,
          "effective_questions": 2,
          "minimum_actions_delta_range": [
            1,
            4
          ],
          "questions": 2,
          "required_chain_depth_delta_range": [
            1,
            1
          ]
        },
        "wide": {
          "all_failed": 1,
          "all_succeeded": 3,
          "effective_questions": 2,
          "minimum_actions_delta_range": [
            -3,
            1
          ],
          "questions": 6,
          "required_chain_depth_delta_range": [
            -1,
            1
          ]
        }
      },
      "seed_observations": {
        "questions": 50,
        "previous_round": 2,
        "current_round": 3,
        "previous_successes": 279,
        "previous_answers": 400,
        "current_successes": 279,
        "current_answers": 400,
        "improved_questions": 12,
        "regressed_questions": 17,
        "unchanged_questions": 21,
        "recovered_from_all_failed": 0,
        "regressed_to_all_failed": 2
      },
      "common_old_task_observations": {
        "questions": 75,
        "previous_round": 2,
        "current_round": 3,
        "previous_successes": 402,
        "previous_answers": 600,
        "current_successes": 407,
        "current_answers": 600,
        "improved_questions": 20,
        "regressed_questions": 22,
        "unchanged_questions": 33,
        "recovered_from_all_failed": 1,
        "regressed_to_all_failed": 4
      },
      "observation_interpretation": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
      "start_utc": "2026-11-02T06:22:35.695521+00:00",
      "end_utc": "2026-11-02T08:10:21.268705+00:00",
      "start_beijing": "2026-11-02 14:22:35",
      "end_beijing": "2026-11-02 16:10:21",
      "full_elapsed_seconds": 6465.573184,
      "stage_sum_seconds": 6465.546611000001,
      "interstage_seconds": 0.026573,
      "recorded_stage_seconds": {
        "plan-round": 4874.832532,
        "train-round": 1590.714079
      },
      "pure_parameter_training_seconds": null,
      "elapsed_scope": "整轮用时包含出题、作答、学习、保存及环节交接；后半段也不是纯模型调整时间。",
      "schema_version": "wide_deep_SCALE50_round3_summary_v1",
      "completed_at_utc": "2026-11-02T08:10:21.268705+00:00",
      "completed_at_beijing": "2026-11-02 16:10:21",
      "started_at_utc": "2026-11-02T06:22:35.695521+00:00",
      "complete_round_seconds": 6465.573184,
      "model_saved": true,
      "main_learning_complete": true,
      "main_learning_updates": 1,
      "new_learning_updates_completed": 1,
      "answers_collected_before_third_learning": true,
      "completed_main_learning_rounds_before_collection": 2,
      "answer_meaning": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
      "final_common30_test_completed": false,
      "source_public_summary": "scale50-round-3-summary.json"
    },
    "completed_all_training_rounds": 21,
    "required_all_training_rounds": 21,
    "second_round_comparison_available": true,
    "latest_completed_OPT": {
      "schema_version": "wide_deep_OPT_round3_summary_v1",
      "arm": "WDT_OPT",
      "round": 3,
      "status": "COMPLETE",
      "observed_at_utc": "2026-11-02T04:48:07.580166+00:00",
      "observed_at_beijing": "2026-11-02 12:48:07",
      "completed_at_utc": "2026-11-02T04:46:39.136471+00:00",
      "completed_at_beijing": "2026-11-02 12:46:39",
      "started_at_utc": "2026-11-02T03:44:15.763341+00:00",
      "questions": 44,
      "answers": 352,
      "seed_questions": 20,
      "seed_answers": 160,
      "seed_successes": 120,
      "previous_round_seed_successes": 125,
      "review_questions": 12,
      "new_questions": 12,
      "effective_questions": 17,
      "all_correct_questions": 20,
      "all_wrong_questions": 7,
      "effective_new_questions": 5,
      "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
      "new_difficulty_by_kind": {
        "deep_down": {
          "all_failed": 0,
          "all_succeeded": 2,
          "effective_questions": 0,
          "minimum_actions_delta_range": [
            -11,
            -2
          ],
          "questions": 2,
          "required_chain_depth_delta_range": [
            -1,
            -1
          ]
        },
        "deep_up": {
          "all_failed": 1,
          "all_succeeded": 1,
          "effective_questions": 3,
          "minimum_actions_delta_range": [
            1,
            13
          ],
          "questions": 5,
          "required_chain_depth_delta_range": [
            1,
            1
          ]
        },
        "wide": {
          "all_failed": 0,
          "all_succeeded": 3,
          "effective_questions": 2,
          "minimum_actions_delta_range": [
            0,
            1
          ],
          "questions": 5,
          "required_chain_depth_delta_range": [
            0,
            0
          ]
        }
      },
      "signal_coverage": {
        "failed_questions": 18,
        "first_answer_failed_questions": 12,
        "independent_student_answers": 0,
        "later_failures_considered": 6,
        "mixed_source_questions": 12,
        "regressed_question_ids": [
          "generated:local_textcraft_248",
          "generated:local_textcraft_280",
          "generated:local_textcraft_462",
          "generated:local_textcraft_328",
          "generated:cluster_wdt_r1_s18",
          "generated:cluster_wdt_r1_s5",
          "generated:cluster_wdt_opt_r2_s14"
        ],
        "source_questions": 32,
        "upward_sources_skipped_without_teacher_proposals": 1
      },
      "main_learning_updates": 1,
      "main_learning_complete": true,
      "model_saved": true,
      "main_learning_reused": false,
      "new_branch_answers": 352,
      "final_policy_id": "5bac08966225f94f9127a524698cf56b33eb41f5ee47ffdbe4bdfd399ed4db45",
      "answers_collected_before_third_learning": true,
      "completed_main_learning_rounds_before_collection": 2,
      "answer_meaning": "第三轮答卷在第三次学习前收集，反映前两轮学习后的表现；本轮模型已保存。",
      "effect_interpretation": "共同20道种子由上一轮125/160变为120/160，当前尚无稳定原题收益；第三次学习的效果待共同30题测试。",
      "final_common30_test_completed": false,
      "complete_round_seconds": 3743.37313,
      "stage_seconds": {
        "plan-round": 2582.338237,
        "train-round": 1161.00745
      },
      "pure_parameter_training_seconds": null,
      "timing_scope": "完整一轮，含出题、作答、学习、保存和阶段切换。",
      "question_results": [
        {
          "task_id": "generated:local_textcraft_188",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_215",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_248",
          "section": "seed",
          "kind": "root",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_273",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_280",
          "section": "seed",
          "kind": "root",
          "successes": 3,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_42",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_448",
          "section": "seed",
          "kind": "root",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_462",
          "section": "seed",
          "kind": "root",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_477",
          "section": "seed",
          "kind": "root",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_480",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_489",
          "section": "seed",
          "kind": "root",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_502",
          "section": "seed",
          "kind": "root",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_517",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_328",
          "section": "seed",
          "kind": "root",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:local_textcraft_357",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_100",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_72",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_45",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_190",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:local_textcraft_355",
          "section": "seed",
          "kind": "root",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_r1_s18",
          "section": "review",
          "kind": "wide",
          "successes": 6,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s4",
          "section": "review",
          "kind": "wide",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_r1_s1",
          "section": "review",
          "kind": "deep_down",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s5",
          "section": "review",
          "kind": "deep_down",
          "successes": 1,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s18",
          "section": "review",
          "kind": "deep_up",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s0",
          "section": "review",
          "kind": "wide",
          "successes": 6,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_r1_s17",
          "section": "review",
          "kind": "deep_down",
          "successes": 6,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s11",
          "section": "review",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s14",
          "section": "review",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s26",
          "section": "review",
          "kind": "deep_down",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s12",
          "section": "review",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r2_s2",
          "section": "review",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s0",
          "section": "new",
          "kind": "wide",
          "successes": 2,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s3",
          "section": "new",
          "kind": "deep_up",
          "successes": 4,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s4",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s15",
          "section": "new",
          "kind": "deep_up",
          "successes": 1,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s18",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s5",
          "section": "new",
          "kind": "deep_down",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s6",
          "section": "new",
          "kind": "wide",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s7",
          "section": "new",
          "kind": "deep_up",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s20",
          "section": "new",
          "kind": "wide",
          "successes": 7,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s19",
          "section": "new",
          "kind": "deep_up",
          "successes": 5,
          "attempts": 8,
          "has_signal": true,
          "included_in_update": true
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s9",
          "section": "new",
          "kind": "deep_up",
          "successes": 0,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        },
        {
          "task_id": "generated:cluster_wdt_opt_r3_s11",
          "section": "new",
          "kind": "deep_down",
          "successes": 8,
          "attempts": 8,
          "has_signal": false,
          "included_in_update": false
        }
      ],
      "source_records": {
        "CAPTURE.json": {
          "bytes": 56218,
          "sha256": "c8d714da49fb7486e237bdaa098f1ebf44cd83326e782d8e371cb3fd22a8f2ee"
        },
        "complete.json": {
          "bytes": 9013,
          "sha256": "4ac1899cd2b0324106031cb4a65764ed1f173599efa2794ff06ee3b798cd8510"
        },
        "collection.json": {
          "bytes": 12803,
          "sha256": "922c6d8d09c7ee673ebb019058dcf4ca9b7a78413292328fb993fba8236e7a55"
        },
        "plan.json": {
          "bytes": 97862,
          "sha256": "329e33b2eca23c3cbd64ec4c8059a356eb96fe7b82348c7043311e6e61ba9d0a"
        },
        "main.json": {
          "bytes": 8309,
          "sha256": "0c127a4cef51b8b980f146cd750fdaa55a93c09672b1396dfb0cc95591bb31e5"
        },
        "main_model/COMPLETE.json": {
          "bytes": 337,
          "sha256": "a554c4e4fa5036b7cd2bdf652584005eab875a59947c9c7276e1c94da806b771"
        },
        "timing/WDT_OPT_plan-round_3_attempt_01.started.json": {
          "bytes": 1126,
          "sha256": "bc14a071791f28522102c815cce1a0c7142d1a6f5996736461dc6145ca9f5e3c"
        },
        "timing/WDT_OPT_train-round_3_attempt_01.exit.json": {
          "bytes": 178,
          "sha256": "197014fad7e059c45549f2f94905a4c2332dcc29e400a688d6e8cacda4bd9e70"
        }
      }
    },
    "OPT_training_complete": true,
    "active_stage_basis": "原会话17:29交接：WT、DT各84/90待补测，WDT已存44/90并继续测试。",
    "REPAIR_training_complete": true,
    "next_revision_candidate": "next_revision_candidate_v7",
    "next_revision_deployed": false,
    "observed_time_precision": "minute",
    "all_main_training_complete": true,
    "main_training_completed_at_utc": "2026-11-02T08:10:21.268705+00:00",
    "main_training_completed_at_beijing": "2026-11-02 16:10:21",
    "SCALE50_training_complete": true,
    "main_test_models_completed": 1,
    "completed_test_models": [
      "N"
    ],
    "completed_test_result_count": 1,
    "base_test_result": {
      "label": "N",
      "complete": true,
      "summary_matches_scope": true,
      "questions": 30,
      "answers": 90,
      "sampled_successes": 27,
      "sampled_attempts": 60,
      "sampled_rate": 0.45,
      "greedy_successes": 15,
      "greedy_attempts": 30,
      "greedy_rate": 0.5
    },
    "auxiliary_test_observed_at_beijing": "2026-11-02 16:37:20",
    "main_test_answers_count_basis": "已完成基础模型90份 + WT已存84份 + DT已存84份 + WDT已存44份。",
    "test_interruptions": [
      {
        "label": "WT",
        "status": "INCOMPLETE_NEEDS_RESUME",
        "answers_saved": 84,
        "answers_required": 90,
        "answers_missing": 6,
        "complete": false,
        "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
        "input_length_limit": 4096,
        "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
        "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
        "recovery_owner": "experiment_main_session",
        "previous_public_record": "wt-test-interruption-1714.json"
      },
      {
        "label": "DT",
        "status": "INCOMPLETE_NEEDS_RESUME",
        "answers_saved": 84,
        "answers_required": 90,
        "answers_missing": 6,
        "complete": false,
        "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
        "input_length_limit": 4096,
        "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
        "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
        "recovery_owner": "experiment_main_session"
      }
    ],
    "completion_estimate_available": false,
    "saved_training_models_unchanged": true
  },
  "observed_at_utc": "2026-11-02T09:29+00:00",
  "observed_at_beijing": "2026-11-02 17:29",
  "page_is_a_live_monitor": false,
  "original_actual_questions_per_round": [
    31,
    40,
    44
  ],
  "completed_original_learning_stages": [
    {
      "arm": "R",
      "round": 2,
      "questions": 40,
      "answers": 320,
      "effective_questions": 16,
      "seed_successes": 120,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "743c00359419df1fa25555540b9039ec8a29ce47f9b63e426eab7381a340e90b"
    },
    {
      "arm": "R",
      "round": 1,
      "questions": 31,
      "answers": 248,
      "effective_questions": 13,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "f3367dbb2b02e3f8c3632c48fc62dcc72c606addb5454e0ef657e5ce65e1d8f8"
    },
    {
      "arm": "WDT",
      "round": 2,
      "questions": 40,
      "answers": 320,
      "effective_questions": 19,
      "seed_successes": 125,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "f1dd916cad49985cad9da01681c71709f12167da36a5187447fcf49ae6118559"
    },
    {
      "arm": "WDT",
      "round": 1,
      "questions": 31,
      "answers": 248,
      "effective_questions": 14,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "d09243aeffd2fb6648963d4d6858622589b46d1e5ed7783f4ad679dfbecc8e9f"
    },
    {
      "arm": "DT",
      "round": 2,
      "questions": 40,
      "answers": 320,
      "effective_questions": 25,
      "seed_successes": 122,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "17f3ed0d0c5ecfaadf78f3992d147b0fc856a8e54c8890ba3d4152d23f1297a8"
    },
    {
      "arm": "DT",
      "round": 1,
      "questions": 31,
      "answers": 248,
      "effective_questions": 16,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "435a8905877d291cabb54cdc8822a4089b3ab868483f397f1473a02fe7f02305"
    },
    {
      "arm": "WT",
      "round": 3,
      "questions": 44,
      "answers": 352,
      "effective_questions": 19,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "3892e4bc8b516da20051190a92fb16472972b7deafe5ebbf539c6f2f89df10d9"
    },
    {
      "arm": "WT",
      "round": 2,
      "questions": 40,
      "answers": 320,
      "effective_questions": 19,
      "seed_successes": 122,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "d85d5d82a18ba9f945ca200ebc63c3681e74866ecc5b00f6b1465bd9aa0d6f10"
    },
    {
      "arm": "WT",
      "round": 1,
      "questions": 31,
      "answers": 248,
      "effective_questions": 15,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "594d5e7e671b4ce88ba73d9747a03b32c3ec52a60109ff68bcd51c0c2ed7b470"
    },
    {
      "arm": "DT",
      "round": 3,
      "questions": 44,
      "answers": 352,
      "effective_questions": 27,
      "seed_successes": 121,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "fef82ff38a7b463f232c2bba7c9ecee6aeab672c3295e9668d4a2e673c9c6475"
    },
    {
      "arm": "WDT",
      "round": 3,
      "questions": 44,
      "answers": 352,
      "effective_questions": 19,
      "seed_successes": 120,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "8b6b702d8b2b91e09968d53c0901989f900e198cf27d5275f3c9c068fb032cd4"
    },
    {
      "arm": "R",
      "round": 3,
      "questions": 44,
      "answers": 352,
      "effective_questions": 21,
      "seed_successes": 120,
      "seed_answers": 160,
      "whole_round_updates": 1,
      "model_saved": true,
      "policy_id": "bfeeac074b891013fae6155b899f51a7bde61be5a1f0e139818a86121529bc4b"
    }
  ],
  "completed_learning_answer_uses": 3680,
  "completed_learning_rounds": 12,
  "required_original_learning_rounds": 12,
  "current_answers_saved": 44,
  "current_answers_required": 90,
  "current_completed_question_groups": null,
  "first_round_summary": "round-1-summary.json",
  "second_round_summary": "round-2-summary.json",
  "extension_status": "TRAINING_COMPLETE",
  "main_test_answers_planned": 720,
  "auxiliary_test_answers_planned": 180,
  "final_test_results_available": false,
  "source_evidence": {
    "progress.json": {
      "bytes": 109470,
      "sha256": "c3f89d0e9397c5a24719d867628d7614dda1b8b8e7b12eac71fb6e24a9b6d13d"
    },
    "observer.json": {
      "bytes": 2474,
      "sha256": "c2a606f034ce5da41f0610c49acf0c47f711429c1c6c7db9a133a9be61f6422c"
    },
    "release-evidence.json": {
      "bytes": 7286,
      "sha256": "6a05a27b5e29ca26ce920c7181f9a818b6823292ef79c12f2d86bae0aff60ca0"
    },
    "bundle-readback.json": {
      "bytes": 900,
      "sha256": "6b8c5ffa3ea72958f992987890309fa1f2973d98edc0eb2687439490eda029eb"
    },
    "progress-update.json": {
      "bytes": 109469,
      "sha256": "e109831c629f761715d65991c407c12781ed2cc6d9fbfecc85d33bd8678d557c"
    },
    "observer-update.json": {
      "bytes": 2473,
      "sha256": "a742e71211a7a33379cdf2540601791fa6eab9b1aba00d194e4f4c6621a400be"
    },
    "extension-prepared.json": {
      "bytes": 901,
      "sha256": "0f862396701469987c8ba5b59edeb692c6392e2931acc35e0bbf8d47a64f9256"
    },
    "wt-round3-complete.json": {
      "bytes": 9392,
      "sha256": "77474b88d01473b0697059ea8ba83c6759f92d97743591ce09a6d17df2c0e6a6"
    },
    "wt-round3-collection.json": {
      "bytes": 12800,
      "sha256": "35ff48072da075120732318696ce90f90e3bf62edb10901dd219d2406cd178d6"
    },
    "dt-round3-complete.json": {
      "bytes": 9400,
      "sha256": "472271f93398ccb4da62d3887067153fbb800a6a86af538909e80fef252c5f04"
    },
    "dt-round3-collection.json": {
      "bytes": 12791,
      "sha256": "6f2e1a9b7d41432a9c9fd456f5e62dd509f9198c56866e66ee488efa146e884c"
    },
    "completion-boundary-050254.json": {
      "bytes": 633,
      "sha256": "ca73a341d5a4b91b4fec8937677ae6f56d4981b4671c9d59664c0dfd3148a32c"
    },
    "started-RUN_ALLOWED.json": {
      "bytes": 742,
      "sha256": "6928a7e6fe2b9d384a0560bdf85448fcc1c24c23f98d173dbd63bbf930abf320"
    },
    "started-BINDING.json": {
      "bytes": 3137,
      "sha256": "2a1d15e7448b5000e7baa6a48272d211156d3db9fe44083c888f1249bb6102a8"
    },
    "started-GUARD_PID.json": {
      "bytes": 391,
      "sha256": "a207c73bd156f36fb50b34bc133ca1111260cf60bcb0f4e249d100c39138a259"
    },
    "started-MONITOR_PID.json": {
      "bytes": 153,
      "sha256": "6a5be518591a6fb65641a5679948260e08434e4deeed0107ef7358c7c094ca4c"
    },
    "started-TIMEOUT_PID.json": {
      "bytes": 564,
      "sha256": "ca8a2fbd22e9575eac1dbc750872ebe7f23f264b24514df30498dc2695e9a8d8"
    },
    "started-PREPARED.json": {
      "bytes": 5947,
      "sha256": "d5941f79a3f5bcc1ba14d81556b44d357def6a5fef05f33a9f006ecd3e8fe173"
    },
    "started-EXTENSION_DISPATCH_SPEC.json": {
      "bytes": 20596,
      "sha256": "902402c95cf31e5e7dcf5696c4bc568af83381f468ce7bd3d9387840d7c1a006"
    },
    "started-STARTUP_CONFIG_RECOVERY.json": {
      "bytes": 1007,
      "sha256": "14c6ebe831abfe37aaf715975495a090b068a41c6719e784f6ad03546b803ae4"
    },
    "started-LOCAL_OBSERVER_STATUS.json": {
      "bytes": 31161,
      "sha256": "a3668ae62c7e124b240cb0f3dfa021e0fe93ff33d5089852056a4348bc9ac94a"
    },
    "started-CURRENT_PROGRESS.json": {
      "bytes": 38646,
      "sha256": "5b4fc86528c53ceb04630cbad6648bbc69d9fe9b1927e6f5209936db78c54a6c"
    },
    "started-CAPTURE.json": {
      "bytes": 2115,
      "sha256": "44b8fc0b675dabab107797b80fa0f66e631f1300b447bddbb24b641bab7e6a47"
    },
    "round3-WDT-complete.json": {
      "bytes": 9384,
      "sha256": "e07e8bf598ec63632805d025fdbc32b7b9a9f9ad32d922c900f63d032e74895e"
    },
    "round3-WDT-collection.json": {
      "bytes": 12807,
      "sha256": "b8baad8621d8b20339dd1058da5995a44bfe3a54e90679ad8c418106b162206f"
    },
    "round3-WDT-main.json": {
      "bytes": 8185,
      "sha256": "42e19ba95d5daa61a732a51c785904b75665fa09bff2912302f848915231a1cb"
    },
    "round3-WDT-main_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "0ee47c9dc15074291bf08dd7b5378322e68734ca97324fa5ab37918bdebadd9a"
    },
    "round3-WDT-family_history.json": {
      "bytes": 156059,
      "sha256": "4fc77a898ca1c588b1cf36ae5b328171b05a098531062412b24da7ef0b13c111"
    },
    "round3-WDT-CAPTURE.json": {
      "bytes": 1613,
      "sha256": "1940cfe48a90ca643d0b87aa9c13e25a471d5269d1474feaea22480b485e1134"
    },
    "round3-R-complete.json": {
      "bytes": 9381,
      "sha256": "d7d0960812e76d3cb66ad40955f4cc5fe116de6560f6d7fa168a26d38eb5e089"
    },
    "round3-R-collection.json": {
      "bytes": 12801,
      "sha256": "64a5c57a7160f442dd3e8386ae55401dd8b0e443b122a9b519a75644da7d53de"
    },
    "round3-R-main.json": {
      "bytes": 8184,
      "sha256": "0494b16018451fe0e605727874b48d39c412896205984ab2b6a0a8285632e175"
    },
    "round3-R-main_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "757207aa83418d8504edd09cff7b65a244927de322871c356b2ac18012bc22d5"
    },
    "round3-R-family_history.json": {
      "bytes": 137120,
      "sha256": "51a1623d6d5efa59267e6a71e54d3eaab199624ea22be1819389b665aa48a6ee"
    },
    "round3-R-CAPTURE.json": {
      "bytes": 1601,
      "sha256": "498af7a80034d540d87aee9a90a7781eec640fcc84c6b752ed107395b24eab65"
    },
    "budget-after-launch.json": {
      "bytes": 4537,
      "sha256": "2ac871499bd4a6806a9677625efcc0517fd2c9d102ae4dc1d71701bbd92dffc6"
    },
    "budget-after-launch.md": {
      "bytes": 894,
      "sha256": "1a6e4cefb4463a0747f480426f646c1793090072c517146ca3bcae9c631d3aba"
    },
    "efficiency-review.md": {
      "bytes": 5825,
      "sha256": "74d81216053bd94c5fba647415a1cb34052f6c3bb5c34ecb60fa46b7caa0aca7"
    },
    "previous-public-budget.json": {
      "bytes": 4320,
      "sha256": "79257c1b416a04aa52fd4716edbca12ebd53b9ee1a14302c6109529fbd69b53e"
    },
    "previous-public-budget.md": {
      "bytes": 1812,
      "sha256": "217348f5ddaecc7f5dd3b96b96d7fa1442830d21af4ed12a118be7eef951e72f"
    },
    "observer-correction-complete-20261101T222214.json": {
      "bytes": 31345,
      "sha256": "491f55c35296d2e94354b6925823a1e25120af6b4b1378770e6ee3d0c926e27b"
    },
    "repair-r1-completed-CAPTURE.json": {
      "bytes": 23268,
      "sha256": "1f46ca8826fc52ad1b469507aa597e6a5017746af20ddd32cf9f4e563b1a4126"
    },
    "repair-r1-completed-REQUEST.json": {
      "bytes": 2144,
      "sha256": "c754e5dc4db32ff6ff2cca05d8c8f359d05eb612be33328ed07e2504fe29696e"
    },
    "repair-r1-completed-collection.json": {
      "bytes": 9329,
      "sha256": "4c5e1e982c4a5283d254f6ce695e602e99bc9d50df38f61d1067f9dc69214625"
    },
    "repair-r1-completed-complete.json": {
      "bytes": 8049,
      "sha256": "989f702bc40be0d6c63ae645199e0d7c1cd644eb205e27b6ad4a418733a4e0f3"
    },
    "repair-r1-completed-family_history.json": {
      "bytes": 72086,
      "sha256": "0916da65e40fe6e75dfb05b9bd7554f9494f564c2462dddb581f07add74d41f5"
    },
    "repair-r1-completed-main.json": {
      "bytes": 6394,
      "sha256": "28d5aa327a869a295a83a9db1ea4cf65d4d9ff25e18b07573226cc43b06ed9fb"
    },
    "repair-r1-completed-main_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "bb9c3e7bfe89a34d990b80ca8c8954dca3cc774cadf0de3669cd34e895a8fa34"
    },
    "repair-r1-completed-plan.json": {
      "bytes": 41156,
      "sha256": "d877600a54656737cf12a7ac324d2014fb3f6760ce57fd43070e0cea0816ef1b"
    },
    "repair-r1-completed-rejected_repair_details.json": {
      "bytes": 71305,
      "sha256": "ecb689cb7f26c160a928b90d3eceb2418c619e2705f1a2bd16d455fd7e47ec0d"
    },
    "repair-r1-completed-repair.json": {
      "bytes": 951,
      "sha256": "8b6730be444c0ff311bef5b39c9ed886b5740bd8576ed1c02ef00007be19be2d"
    },
    "repair-r1-completed-repair_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "bffe8fd95f99bf78ffd542501a647345704ca22eb55bd0c07985209ce3160b36"
    },
    "repair-r1-completed-repair_summary.json": {
      "bytes": 75847,
      "sha256": "8268f7440eb8b8c692d74c9777b8cf029a866a65ccae9bbdfc44a619982589b2"
    },
    "repair-r1-completed-timing-WDT_REPAIR_train-round_1_attempt_01.exit.json": {
      "bytes": 181,
      "sha256": "4032375f73c4971b150138fe889d778553127836ec4981f32e4694b8fd7a9200"
    },
    "repair-r1-completed-timing-WDT_REPAIR_train-round_1_attempt_01.started.json": {
      "bytes": 1148,
      "sha256": "5ce5432338172c516657a22ecc901c84b4619ff361e0d284a82f9c486848cf67"
    },
    "progress-scale50-r1-complete-080800.json": {
      "bytes": 40486,
      "sha256": "ff40835954f67529563564c6634214cf63b8545830b3f8dd8266aa0a8007fb83"
    },
    "scale50-r1-completed-CAPTURE.json": {
      "bytes": 46707,
      "sha256": "23078f59fe5d597159ae6e7fca88c3f02428d75a4af55eb1b2f56eb997fbfa7a"
    },
    "scale50-r1-completed-complete.json": {
      "bytes": 13857,
      "sha256": "6c13660a6b25dd5b3e2cdfe3f9e1668ef223a21cfd3dc3000c67fd50398af46e"
    },
    "scale50-r1-completed-collection.json": {
      "bytes": 21631,
      "sha256": "b1f1a64a13566b35cfbe1f477ffbf5a34e6fb7f05f6b150ead3816c57a4d2517"
    },
    "scale50-r1-completed-main.json": {
      "bytes": 13146,
      "sha256": "b6c09cf93d3d05fab95332aff0055b1d2e0210fd8ce259a72b54408e5cbc2000"
    },
    "scale50-r1-completed-plan.json": {
      "bytes": 146372,
      "sha256": "c7596e07e4c942f8f1f556d8deb8cf2a97fc04d5510554e77566e55bbe3356c1"
    },
    "scale50-r1-timing.json": {
      "bytes": 2940,
      "sha256": "0b3dbd691adcbe33ee26f742af412f73fde7f40ae0bd204a0c427e5393d3a685"
    },
    "budget-after-scale50-r1.json": {
      "bytes": 19622,
      "sha256": "1e72d5bbee21aae219261644162593e97d5d86bf963f051b2b77aca404b2de8d"
    },
    "budget-after-scale50-r1.md": {
      "bytes": 3134,
      "sha256": "bbffff31ac43499ba5c79623e2a023349f55f3c68f6a934ea55d56f1039333e4"
    },
    "OBSERVATION_122003.json": {
      "bytes": 32440,
      "sha256": "9ac2993c5cac1101293b0fb8d75a6c50cffe9d22ada7016682529abb7c426b00"
    },
    "extension-round2-summary.json": {
      "bytes": 14441,
      "sha256": "6fe98bea611cfa7a7e94c912cf1a966555ef2684c9d071888ff1c46671a634a9"
    },
    "scale50-round2-generation.json": {
      "bytes": 1367166,
      "sha256": "b710a6d99f7f90af62c0fe62af62e5529354b0c5456e415d7e79660b9ef59f39"
    },
    "budget-after-scale50-round2.json": {
      "bytes": 30625,
      "sha256": "1a70b2321e59cf6c36b5142e0ad7ff3dd8452d788605431c61d516359052f29c"
    },
    "OPT_R2_COMPLETED_20261102_CAPTURE02-CAPTURE.json": {
      "bytes": 41850,
      "sha256": "a26b94fde3da906b1ea34543dda655657c43d2664c1479394c0f334e692336f5"
    },
    "REPAIR_R2_COMPLETED_20261102_CAPTURE01-CAPTURE.json": {
      "bytes": 43850,
      "sha256": "d08c7b28c2c8e7f959c8a3b39234c9bcb64dda4a4f2eec6b17ed327d5f989199"
    },
    "SCALE50_R2_COMPLETED_20261102_CAPTURE01-CAPTURE.json": {
      "bytes": 81497,
      "sha256": "91e067ae2740e14e5df137b4df9fbc001f52428b422b65b7df3d1c5be57bd069"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-CAPTURE.json": {
      "bytes": 56218,
      "sha256": "c8d714da49fb7486e237bdaa098f1ebf44cd83326e782d8e371cb3fd22a8f2ee"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-complete.json": {
      "bytes": 9013,
      "sha256": "4ac1899cd2b0324106031cb4a65764ed1f173599efa2794ff06ee3b798cd8510"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-collection.json": {
      "bytes": 12803,
      "sha256": "922c6d8d09c7ee673ebb019058dcf4ca9b7a78413292328fb993fba8236e7a55"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-plan.json": {
      "bytes": 97862,
      "sha256": "329e33b2eca23c3cbd64ec4c8059a356eb96fe7b82348c7043311e6e61ba9d0a"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-main.json": {
      "bytes": 8309,
      "sha256": "0c127a4cef51b8b980f146cd750fdaa55a93c09672b1396dfb0cc95591bb31e5"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-main_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "a554c4e4fa5036b7cd2bdf652584005eab875a59947c9c7276e1c94da806b771"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_OPT_plan-round_3_attempt_01.started.json": {
      "bytes": 1126,
      "sha256": "bc14a071791f28522102c815cce1a0c7142d1a6f5996736461dc6145ca9f5e3c"
    },
    "OPT_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_OPT_train-round_3_attempt_01.exit.json": {
      "bytes": 178,
      "sha256": "197014fad7e059c45549f2f94905a4c2332dcc29e400a688d6e8cacda4bd9e70"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-CAPTURE.json": {
      "bytes": 58292,
      "sha256": "481c8837533874f3fcc9d252a3779276d4967bba13aaf908273090373464d3c0"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-complete.json": {
      "bytes": 9988,
      "sha256": "94889594a144983fd0a0bd53ce66a79f38c72aeb7b8218d4be6868492535af1e"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-collection.json": {
      "bytes": 12803,
      "sha256": "05ccda0149e1ca38cabe05bb2db12793d5cda60a5853ef2ecaa99a684e720b6f"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-plan.json": {
      "bytes": 101855,
      "sha256": "cf4a8aac703816624a1ad4bc09cc51656228a6ff920b5947ab9f1286b220e0c8"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-main.json": {
      "bytes": 8330,
      "sha256": "80ea9fbc1874462bb70cd08e5b134e9ead463148db9d672c67b66e6020b6a759"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-repair.json": {
      "bytes": 952,
      "sha256": "a9aac9f467d3dd314e804ea9a0205150d5607d8ca7928fd4b9ec097e4a6008c5"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-repair_summary.json": {
      "bytes": 116040,
      "sha256": "3c50f288bb48e2ec1440abfa4794cfc601f79c6ba574a786780a675d4dbab041"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-main_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "9af78ea8fced779502f596ff16b3076f12dc05de0324b1110ac2a350e4649d99"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-repair_model-COMPLETE.json": {
      "bytes": 337,
      "sha256": "938db3362a51c2379bf21664056a7281fe2cfe67a92f2a47b8dad82bd397ea49"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_REPAIR_plan-round_3_attempt_01.started.json": {
      "bytes": 1141,
      "sha256": "b9c492335e3fbf15cb8a3a29143d5b1d0416666fd571fce252f7f02e72208b5c"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_REPAIR_plan-round_3_attempt_01.exit.json": {
      "bytes": 180,
      "sha256": "5dc4349ad5667c350da7433bf1d232e41b7b24498a636f6de6f52b4c60d5d19f"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_REPAIR_train-round_3_attempt_01.started.json": {
      "bytes": 1148,
      "sha256": "76bf2a00a9f18084a8d6b25eb69acf9225e24c4cf8e08aaaffe85c9dc412edc5"
    },
    "REPAIR_R3_COMPLETED_20261102_CAPTURE01-timing-WDT_REPAIR_train-round_3_attempt_01.exit.json": {
      "bytes": 181,
      "sha256": "e6fe2939a08c4705ad4a143a9e4599d52f6d6b8a92c86fae887284913c6153e1"
    },
    "REPAIR_R3_TRAINING_20261102.json": {
      "bytes": 8575,
      "sha256": "559e0778b28fff5ece438ae2b57f577dfd0d50c06e2ffe4829c427fd68a9e80b"
    },
    "fixed_observation_20261102_142820": {
      "bytes": 34379,
      "sha256": "10accc79c815d530ea2d31555981f4d9f50dfacbaddbe498c8add4ce346c1b16"
    },
    "scale50-round-3-summary.json": {
      "bytes": 7354,
      "sha256": "8b5f92783de33d95a60b99eab5952fc1ebf134eea3af07e4de7bd3d68dab97d0"
    },
    "base-test-summary.json": {
      "bytes": 10015,
      "sha256": "f8e2653d5e081a3804b0c1b5b5cd208d3fa2c70bb6c4c1624ac07a56486109f5"
    },
    "test-progress-164452.json": {
      "bytes": 749,
      "sha256": "4df8bbaf7ccb2cc4f91ffea134e7881240b0a36e474a64a3bfd0213b94e5aa9d"
    },
    "wt-test-interruption-1714.json": {
      "bytes": 1746,
      "sha256": "6896cc2d47d936e7b217ad7fe5325b2b31b8995e37b33a6d0bdad106a5a97955"
    },
    "test-progress-1729.json": {
      "bytes": 2158,
      "sha256": "07bf73ffd4512a1e3780e3a0137aee7cb0e5f4338ab0d93ece4f18032fa08008"
    }
  },
  "WT_third_round": {
    "questions": 44,
    "answers": 352,
    "mixed_questions": 19,
    "all_correct_questions": 15,
    "all_wrong_questions": 10,
    "main_learning_complete": true,
    "model_saved": true,
    "collected_answers_precede_third_round_learning": true,
    "same_40_old_questions": {
      "before_round2_learning_successes": 196,
      "after_round2_learning_successes": 194,
      "answers_each": 320
    },
    "seed_successes": 121,
    "seed_answers": 160,
    "completed_learning_rounds_before_collection": 2
  },
  "current_answer_count_available_in_snapshot": true,
  "DT_third_round": {
    "questions": 44,
    "answers": 352,
    "mixed_questions": 27,
    "all_correct_questions": 15,
    "all_wrong_questions": 2,
    "seed_successes": 121,
    "seed_answers": 160,
    "main_learning_complete": true,
    "model_saved": true,
    "collected_answers_precede_third_round_learning": true,
    "completed_learning_rounds_before_collection": 2
  },
  "completed_learning_rounds_scope": "original_four_groups",
  "shared_initial_answer_uses_avoided": 480,
  "completed_physical_original_training_answers": 3200,
  "original_training_complete": true,
  "current_answers_origin": "main_common_test_WDT",
  "current_new_main_answers_saved": 0,
  "extension_controller_started": true,
  "new_learning_updates_completed": 10,
  "completed_new_learning_stages": [
    {
      "arm": "WDT_REPAIR",
      "round": 1,
      "phase": "correction",
      "status": "COMPLETE",
      "imported_main_learning": true,
      "reused_main_questions": 31,
      "reused_main_answers": 248,
      "new_learning_updates_completed": 1,
      "updated": true,
      "model_saved": true,
      "proposal_count": 39,
      "accepted_proposals": 34,
      "repaired_original_targets": 12,
      "selected_learning_actions": 49,
      "selected_correction_tokens": 710,
      "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
      "original_target_cases": [
        {
          "task_id": "generated:local_textcraft_489",
          "goal": "minecraft:red_banner",
          "proposals": 3,
          "accepted_proposals": 3,
          "selected_learning_actions": 4
        },
        {
          "task_id": "generated:local_textcraft_477",
          "goal": "minecraft:bookshelf",
          "proposals": 3,
          "accepted_proposals": 0,
          "selected_learning_actions": 0
        }
      ],
      "stage_seconds": 864.250016,
      "stage_time_scope": "纠错收集、学习、保存和准备在内的完整首轮阶段。",
      "student_independent_success_after_learning_verified": false,
      "final_policy_id": "5ad18b4b0b35fb42026592d3650659ad39bdc6e65efd8d03a03bab7dfe1e5d14"
    },
    {
      "arm": "WDT_SCALE50",
      "round": 1,
      "phase": "practice",
      "status": "COMPLETE",
      "questions": 75,
      "answers": 600,
      "effective_questions": 37,
      "new_learning_updates_completed": 1,
      "updated": true,
      "model_saved": true,
      "imported_main_learning": false,
      "final_policy_id": "fab4264a5e01094a423b60df09b4fa894608e2e7c3be889572cd5eeb57784da0"
    },
    {
      "arm": "WDT_OPT",
      "round": 2,
      "phase": "practice",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1
    },
    {
      "arm": "WDT_REPAIR",
      "round": 2,
      "phase": "practice",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1
    },
    {
      "arm": "WDT_REPAIR",
      "round": 2,
      "phase": "correction",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1,
      "selected_learning_actions": 65
    },
    {
      "arm": "WDT_SCALE50",
      "round": 2,
      "phase": "practice",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1
    },
    {
      "arm": "WDT_OPT",
      "round": 3,
      "phase": "practice",
      "status": "COMPLETE",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1,
      "questions": 44,
      "answers": 352,
      "effective_questions": 17
    },
    {
      "arm": "WDT_REPAIR",
      "round": 3,
      "phase": "practice",
      "status": "COMPLETE",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1,
      "questions": 44,
      "answers": 352,
      "effective_questions": 18,
      "final_policy_id": "f445a3e252ec14df358c1fc37bb26fd7c0756cce8034ad607926b0642b7330a7"
    },
    {
      "arm": "WDT_REPAIR",
      "round": 3,
      "phase": "correction",
      "status": "COMPLETE",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1,
      "proposal_count": 69,
      "original_goal_completed_proposals": 47,
      "accepted_proposals": 46,
      "original_target_count": 23,
      "repaired_original_targets": 18,
      "selected_learning_actions": 73,
      "selected_correction_tokens": 1055,
      "student_independent_success_after_learning_verified": false,
      "final_policy_id": "aeca99455724c3b1855ef6b0ed30ea2fce7dd20dfd01f7a907d386b0c2667ce3"
    },
    {
      "arm": "WDT_SCALE50",
      "round": 3,
      "phase": "practice",
      "status": "COMPLETE",
      "updated": true,
      "model_saved": true,
      "new_learning_updates_completed": 1,
      "questions": 84,
      "answers": 672,
      "effective_questions": 40,
      "source_public_summary": "scale50-round-3-summary.json"
    }
  ],
  "main_test_answers_saved": 302,
  "auxiliary_test_answers_saved": 0,
  "WDT_third_round": {
    "questions": 44,
    "answers": 352,
    "mixed_questions": 19,
    "all_correct_questions": 19,
    "all_wrong_questions": 6,
    "seed_successes": 120,
    "seed_answers": 160,
    "main_learning_complete": true,
    "model_saved": true,
    "final_policy_id": "8b6b702d8b2b91e09968d53c0901989f900e198cf27d5275f3c9c068fb032cd4",
    "collected_answers_precede_third_round_learning": true,
    "completed_learning_rounds_before_collection": 2,
    "new_question_results": [
      {
        "task_id": "generated:cluster_wdt_r3_s7",
        "kind": "deep_up",
        "successes": 0,
        "answers": 8
      },
      {
        "task_id": "generated:cluster_wdt_r3_s13",
        "kind": "deep_up",
        "successes": 0,
        "answers": 8
      },
      {
        "task_id": "generated:cluster_wdt_r3_s19",
        "kind": "deep_up",
        "successes": 0,
        "answers": 8
      },
      {
        "task_id": "generated:cluster_wdt_r3_s3",
        "kind": "deep_down",
        "successes": 8,
        "answers": 8
      }
    ],
    "new_questions_with_mixed_answers": 0
  },
  "R_third_round": {
    "questions": 44,
    "answers": 352,
    "mixed_questions": 21,
    "all_correct_questions": 18,
    "all_wrong_questions": 5,
    "seed_successes": 120,
    "seed_answers": 160,
    "main_learning_complete": true,
    "model_saved": true,
    "final_policy_id": "bfeeac074b891013fae6155b899f51a7bde61be5a1f0e139818a86121529bc4b",
    "collected_answers_precede_third_round_learning": true,
    "completed_learning_rounds_before_collection": 2,
    "seed_successes_by_round": [
      121,
      120,
      120
    ]
  },
  "extension_progress": [
    {
      "label": "WDT_OPT",
      "display_label": "OPT",
      "seed_questions": 20,
      "round_question_caps": [
        40,
        60,
        60
      ],
      "target_questions": [
        31,
        40,
        44
      ],
      "first_main_round_source": "original_WDT_round_1",
      "new_main_rounds": [
        2,
        3
      ],
      "new_main_answers": 672,
      "main_updates_per_round": 1,
      "repair": false,
      "review_nonseed_fraction_max": 0.5,
      "training_started": true,
      "independent_tasks_from_round": 2,
      "status": "TRAINING_COMPLETE",
      "status_label": "三轮已完成并保存",
      "current_phase": "waiting_final_test",
      "new_learning_started": true,
      "new_learning_updates_completed": 2,
      "imported_main_learning": true,
      "reused_first_round_questions": 31,
      "reused_first_round_answers": 248,
      "new_main_answers_saved": 672,
      "round_progress": [
        {
          "round": 1,
          "status": "COMPLETE",
          "imported_main_learning": true,
          "new_learning_updates_completed": 0,
          "questions": 31,
          "answers_saved": 248,
          "effective_questions": 14,
          "model_saved": true
        },
        {
          "round": 2,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 1,
          "questions": 40,
          "answers_saved": 320,
          "effective_questions": 17,
          "model_saved": true
        },
        {
          "round": 3,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 1,
          "questions": 44,
          "answers_saved": 352,
          "effective_questions": 17,
          "model_saved": true
        }
      ],
      "active_round": null,
      "current_answers_required": 0,
      "current_answers_upper_bound": 0,
      "current_completed_question_groups": 44,
      "second_round_result": {
        "arm": "WDT_OPT",
        "round": 2,
        "status": "COMPLETE",
        "questions": 40,
        "answers": 320,
        "seed_questions": 20,
        "seed_answers": 160,
        "seed_successes": 125,
        "review_questions": 10,
        "new_questions": 10,
        "effective_questions": 17,
        "final_policy_id": "69cec642ec57251107eebc71b4920feb9db9e5c0e8a660fe9c30f070019e92e9",
        "completed_at_utc": "2026-11-02T00:59:16.661732+00:00",
        "completed_at_beijing": "2026-11-02 08:59:16",
        "started_at_utc": "2026-11-02T00:05:42.761413+00:00",
        "complete_round_seconds": 3213.900319,
        "stage_seconds": {
          "plan-round": 2116.67438,
          "train-round": 1097.208328
        },
        "pure_parameter_training_seconds": null,
        "all_correct_questions": 16,
        "all_wrong_questions": 7,
        "effective_new_questions": 4,
        "new_difficulty_by_kind": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 0,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              -17,
              -17
            ],
            "questions": 1,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 3,
            "all_succeeded": 0,
            "effective_questions": 2,
            "minimum_actions_delta_range": [
              1,
              8
            ],
            "questions": 5,
            "required_chain_depth_delta_range": [
              1,
              2
            ]
          },
          "wide": {
            "all_failed": 1,
            "all_succeeded": 2,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              0,
              1
            ],
            "questions": 4,
            "required_chain_depth_delta_range": [
              0,
              1
            ]
          }
        },
        "signal_coverage": {
          "failed_questions": 16,
          "first_answer_failed_questions": 10,
          "independent_student_answers": 0,
          "later_failures_considered": 6,
          "mixed_source_questions": 13,
          "regressed_question_ids": [
            "generated:local_textcraft_357",
            "generated:local_textcraft_190",
            "generated:cluster_wdt_r1_s4",
            "generated:cluster_wdt_r1_s18"
          ],
          "source_questions": 30,
          "upward_sources_skipped_without_teacher_proposals": 1
        },
        "reserved_new_question_positions": 10,
        "main_learning_updates": 1,
        "main_learning_complete": true,
        "model_saved": true,
        "repair": null,
        "answers_collected_before_second_learning": true,
        "completed_main_learning_rounds_before_collection": 1,
        "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
        "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
        "source_records": {
          "CAPTURE.json": {
            "bytes": 41850,
            "sha256": "a26b94fde3da906b1ea34543dda655657c43d2664c1479394c0f334e692336f5"
          },
          "complete.json": {
            "bytes": 8406,
            "sha256": "6890c386929017b6f72448edd03bd29afe02aa2f451945d58450a79aaa82dc0e"
          },
          "collection.json": {
            "bytes": 11663,
            "sha256": "21b5604ffc73ff9cb319affda280e0baf6ef9b65710b7f98abe4db37f17671a2"
          },
          "plan.json": {
            "bytes": 86396,
            "sha256": "e34be224d2fad9f691ce6ae821d1e5952031e7072705e45620d2f7fd2fb20002"
          },
          "main.json": {
            "bytes": 7702,
            "sha256": "3f6a9c3729731485b372e375f74a7d13464366c0e11400d6e4fcf909a2005147"
          }
        }
      },
      "third_round_result": {
        "schema_version": "wide_deep_OPT_round3_summary_v1",
        "arm": "WDT_OPT",
        "round": 3,
        "status": "COMPLETE",
        "observed_at_utc": "2026-11-02T04:48:07.580166+00:00",
        "observed_at_beijing": "2026-11-02 12:48:07",
        "completed_at_utc": "2026-11-02T04:46:39.136471+00:00",
        "completed_at_beijing": "2026-11-02 12:46:39",
        "started_at_utc": "2026-11-02T03:44:15.763341+00:00",
        "questions": 44,
        "answers": 352,
        "seed_questions": 20,
        "seed_answers": 160,
        "seed_successes": 120,
        "previous_round_seed_successes": 125,
        "review_questions": 12,
        "new_questions": 12,
        "effective_questions": 17,
        "all_correct_questions": 20,
        "all_wrong_questions": 7,
        "effective_new_questions": 5,
        "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
        "new_difficulty_by_kind": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 2,
            "effective_questions": 0,
            "minimum_actions_delta_range": [
              -11,
              -2
            ],
            "questions": 2,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 1,
            "all_succeeded": 1,
            "effective_questions": 3,
            "minimum_actions_delta_range": [
              1,
              13
            ],
            "questions": 5,
            "required_chain_depth_delta_range": [
              1,
              1
            ]
          },
          "wide": {
            "all_failed": 0,
            "all_succeeded": 3,
            "effective_questions": 2,
            "minimum_actions_delta_range": [
              0,
              1
            ],
            "questions": 5,
            "required_chain_depth_delta_range": [
              0,
              0
            ]
          }
        },
        "signal_coverage": {
          "failed_questions": 18,
          "first_answer_failed_questions": 12,
          "independent_student_answers": 0,
          "later_failures_considered": 6,
          "mixed_source_questions": 12,
          "regressed_question_ids": [
            "generated:local_textcraft_248",
            "generated:local_textcraft_280",
            "generated:local_textcraft_462",
            "generated:local_textcraft_328",
            "generated:cluster_wdt_r1_s18",
            "generated:cluster_wdt_r1_s5",
            "generated:cluster_wdt_opt_r2_s14"
          ],
          "source_questions": 32,
          "upward_sources_skipped_without_teacher_proposals": 1
        },
        "main_learning_updates": 1,
        "main_learning_complete": true,
        "model_saved": true,
        "main_learning_reused": false,
        "new_branch_answers": 352,
        "final_policy_id": "5bac08966225f94f9127a524698cf56b33eb41f5ee47ffdbe4bdfd399ed4db45",
        "answers_collected_before_third_learning": true,
        "completed_main_learning_rounds_before_collection": 2,
        "answer_meaning": "第三轮答卷在第三次学习前收集，反映前两轮学习后的表现；本轮模型已保存。",
        "effect_interpretation": "共同20道种子由上一轮125/160变为120/160，当前尚无稳定原题收益；第三次学习的效果待共同30题测试。",
        "final_common30_test_completed": false,
        "complete_round_seconds": 3743.37313,
        "stage_seconds": {
          "plan-round": 2582.338237,
          "train-round": 1161.00745
        },
        "pure_parameter_training_seconds": null,
        "timing_scope": "完整一轮，含出题、作答、学习、保存和阶段切换。",
        "question_results": [
          {
            "task_id": "generated:local_textcraft_188",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_215",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_248",
            "section": "seed",
            "kind": "root",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_273",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_280",
            "section": "seed",
            "kind": "root",
            "successes": 3,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_42",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_448",
            "section": "seed",
            "kind": "root",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_462",
            "section": "seed",
            "kind": "root",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_477",
            "section": "seed",
            "kind": "root",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_480",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_489",
            "section": "seed",
            "kind": "root",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_502",
            "section": "seed",
            "kind": "root",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_517",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_328",
            "section": "seed",
            "kind": "root",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_357",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_100",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_72",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_45",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_190",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_355",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_r1_s18",
            "section": "review",
            "kind": "wide",
            "successes": 6,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s4",
            "section": "review",
            "kind": "wide",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_r1_s1",
            "section": "review",
            "kind": "deep_down",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s5",
            "section": "review",
            "kind": "deep_down",
            "successes": 1,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s18",
            "section": "review",
            "kind": "deep_up",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s0",
            "section": "review",
            "kind": "wide",
            "successes": 6,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s17",
            "section": "review",
            "kind": "deep_down",
            "successes": 6,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s11",
            "section": "review",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s14",
            "section": "review",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s26",
            "section": "review",
            "kind": "deep_down",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s12",
            "section": "review",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r2_s2",
            "section": "review",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s0",
            "section": "new",
            "kind": "wide",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s3",
            "section": "new",
            "kind": "deep_up",
            "successes": 4,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s4",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s15",
            "section": "new",
            "kind": "deep_up",
            "successes": 1,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s18",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s5",
            "section": "new",
            "kind": "deep_down",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s6",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s7",
            "section": "new",
            "kind": "deep_up",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s20",
            "section": "new",
            "kind": "wide",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s19",
            "section": "new",
            "kind": "deep_up",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s9",
            "section": "new",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_opt_r3_s11",
            "section": "new",
            "kind": "deep_down",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          }
        ],
        "source_records": {
          "CAPTURE.json": {
            "bytes": 56218,
            "sha256": "c8d714da49fb7486e237bdaa098f1ebf44cd83326e782d8e371cb3fd22a8f2ee"
          },
          "complete.json": {
            "bytes": 9013,
            "sha256": "4ac1899cd2b0324106031cb4a65764ed1f173599efa2794ff06ee3b798cd8510"
          },
          "collection.json": {
            "bytes": 12803,
            "sha256": "922c6d8d09c7ee673ebb019058dcf4ca9b7a78413292328fb993fba8236e7a55"
          },
          "plan.json": {
            "bytes": 97862,
            "sha256": "329e33b2eca23c3cbd64ec4c8059a356eb96fe7b82348c7043311e6e61ba9d0a"
          },
          "main.json": {
            "bytes": 8309,
            "sha256": "0c127a4cef51b8b980f146cd750fdaa55a93c09672b1396dfb0cc95591bb31e5"
          },
          "main_model/COMPLETE.json": {
            "bytes": 337,
            "sha256": "a554c4e4fa5036b7cd2bdf652584005eab875a59947c9c7276e1c94da806b771"
          },
          "timing/WDT_OPT_plan-round_3_attempt_01.started.json": {
            "bytes": 1126,
            "sha256": "bc14a071791f28522102c815cce1a0c7142d1a6f5996736461dc6145ca9f5e3c"
          },
          "timing/WDT_OPT_train-round_3_attempt_01.exit.json": {
            "bytes": 178,
            "sha256": "197014fad7e059c45549f2f94905a4c2332dcc29e400a688d6e8cacda4bd9e70"
          }
        }
      },
      "training_complete": true
    },
    {
      "label": "WDT_REPAIR",
      "display_label": "REPAIR",
      "seed_questions": 20,
      "round_question_caps": [
        40,
        60,
        60
      ],
      "target_questions": [
        31,
        40,
        44
      ],
      "first_main_round_source": "original_WDT_round_1",
      "new_main_rounds": [
        2,
        3
      ],
      "new_main_answers": 672,
      "main_updates_per_round": 1,
      "repair": true,
      "repair_rounds": [
        1,
        2,
        3
      ],
      "repair_updates_per_round_max": 1,
      "review_nonseed_fraction_max": 0.5,
      "training_started": true,
      "independent_tasks_from_round": 2,
      "status": "TRAINING_COMPLETE",
      "status_label": "三轮主练和纠错已完成并保存",
      "current_phase": "waiting_final_test",
      "new_learning_started": true,
      "new_learning_updates_completed": 5,
      "imported_main_learning": true,
      "reused_first_round_questions": 31,
      "reused_first_round_answers": 248,
      "new_main_answers_saved": 672,
      "round_progress": [
        {
          "round": 1,
          "status": "COMPLETE",
          "imported_main_learning": true,
          "new_learning_updates_completed": 1,
          "questions": 31,
          "answers_saved": 248,
          "effective_questions": 14,
          "model_saved": true
        },
        {
          "round": 2,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 2,
          "questions": 40,
          "answers_saved": 320,
          "effective_questions": 15,
          "model_saved": true
        },
        {
          "round": 3,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 2,
          "questions": 44,
          "answers_saved": 352,
          "effective_questions": 18,
          "model_saved": true,
          "main_learning_complete": true,
          "repair_learning_complete": true
        }
      ],
      "repair_progress": {
        "round": 2,
        "phase": "correction",
        "status": "COMPLETE",
        "collection_saved": true,
        "learning_started": true,
        "model_saved": true,
        "complete": true,
        "selected_learning_actions": 65,
        "selected_correction_tokens": 862,
        "new_learning_updates_completed": 1,
        "proposal_count": 57,
        "accepted_proposals": 41,
        "repaired_original_targets": 16,
        "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。"
      },
      "second_round_result": {
        "arm": "WDT_REPAIR",
        "round": 2,
        "status": "COMPLETE",
        "questions": 40,
        "answers": 320,
        "seed_questions": 20,
        "seed_answers": 160,
        "seed_successes": 125,
        "review_questions": 10,
        "new_questions": 10,
        "effective_questions": 15,
        "final_policy_id": "64bba71af88c1427cc676e23f26a83f3f6f70ead23698dae58c93bd283d64fe7",
        "completed_at_utc": "2026-11-02T02:13:21.055887+00:00",
        "completed_at_beijing": "2026-11-02 10:13:21",
        "started_at_utc": "2026-11-02T00:59:16.683471+00:00",
        "complete_round_seconds": 4444.372416,
        "stage_seconds": {
          "plan-round": 1982.335677,
          "train-round": 2462.02242
        },
        "pure_parameter_training_seconds": null,
        "all_correct_questions": 17,
        "all_wrong_questions": 8,
        "effective_new_questions": 3,
        "new_difficulty_by_kind": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 1,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              -17,
              -17
            ],
            "questions": 2,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 3,
            "all_succeeded": 0,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              1,
              4
            ],
            "questions": 4,
            "required_chain_depth_delta_range": [
              1,
              1
            ]
          },
          "wide": {
            "all_failed": 1,
            "all_succeeded": 2,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              0,
              1
            ],
            "questions": 4,
            "required_chain_depth_delta_range": [
              0,
              0
            ]
          }
        },
        "signal_coverage": {
          "failed_questions": 16,
          "first_answer_failed_questions": 10,
          "independent_student_answers": 0,
          "later_failures_considered": 6,
          "mixed_source_questions": 12,
          "regressed_question_ids": [
            "generated:local_textcraft_280",
            "generated:local_textcraft_448",
            "generated:local_textcraft_357",
            "generated:cluster_wdt_r1_s4",
            "generated:cluster_wdt_r1_s5",
            "generated:cluster_wdt_r1_s18"
          ],
          "source_questions": 30,
          "upward_sources_skipped_without_teacher_proposals": 1
        },
        "reserved_new_question_positions": 10,
        "main_learning_updates": 1,
        "main_learning_complete": true,
        "model_saved": true,
        "repair": {
          "accepted_by_mode": {
            "prevention": 28,
            "recovery": 13
          },
          "accepted_proposals": 41,
          "optimizer_steps": 1,
          "proposal_count": 57,
          "rejection_reasons": {
            "NO_SUCCESSFUL_RELEVANT_CORRECTION": 1,
            "SUFFIX_DID_NOT_COMPLETE_ORIGINAL_TASK": 15
          },
          "repaired_target_ids": [
            "generated:cluster_wdt_r1_s0",
            "generated:cluster_wdt_r1_s18",
            "generated:cluster_wdt_r1_s4",
            "generated:cluster_wdt_repair_r2_s12",
            "generated:cluster_wdt_repair_r2_s2",
            "generated:cluster_wdt_repair_r2_s3",
            "generated:cluster_wdt_repair_r2_s5",
            "generated:local_textcraft_215",
            "generated:local_textcraft_248",
            "generated:local_textcraft_280",
            "generated:local_textcraft_357",
            "generated:local_textcraft_448",
            "generated:local_textcraft_462",
            "generated:local_textcraft_480",
            "generated:local_textcraft_489",
            "generated:local_textcraft_502"
          ],
          "selected_actions": 65
        },
        "answers_collected_before_second_learning": true,
        "completed_main_learning_rounds_before_collection": 1,
        "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
        "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
        "source_records": {
          "CAPTURE.json": {
            "bytes": 43850,
            "sha256": "d08c7b28c2c8e7f959c8a3b39234c9bcb64dda4a4f2eec6b17ed327d5f989199"
          },
          "complete.json": {
            "bytes": 9348,
            "sha256": "2bad3cd116d11f963246df352be563ae4a5c70501957b5fc37def83e0d193964"
          },
          "collection.json": {
            "bytes": 11662,
            "sha256": "3dae285d9255d10c86bb599c9e832f4d0f0d7d5733506d056f8631789fdb17c5"
          },
          "plan.json": {
            "bytes": 90906,
            "sha256": "ef8012fc618a629be84d4c8ac1a1dbb6990abdc57ad29091d548e1cfad38bcbe"
          },
          "main.json": {
            "bytes": 7690,
            "sha256": "a48a7b6a0280e17a2ccb51f341939339875e2bbab3049bb5482235e302b8a9fb"
          }
        }
      },
      "active_round": null,
      "current_completed_question_groups": 44,
      "current_answers_required": 0,
      "current_answers_upper_bound": 0,
      "current_answer_count_available_in_snapshot": true,
      "third_round_result": {
        "schema_version": "wide_deep_REPAIR_round3_summary_v1",
        "arm": "WDT_REPAIR",
        "round": 3,
        "status": "COMPLETE",
        "observed_at_utc": "2026-11-02T06:28:20.285682+00:00",
        "observed_at_beijing": "2026-11-02 14:28:20",
        "started_at_utc": "2026-11-02T04:46:39.165028+00:00",
        "completed_at_utc": "2026-11-02T06:22:35.663424+00:00",
        "completed_at_beijing": "2026-11-02 14:22:35",
        "questions": 44,
        "answers": 352,
        "seed_questions": 20,
        "seed_answers": 160,
        "seed_successes": 118,
        "previous_round_seed_successes": 125,
        "common_old_questions": 32,
        "common_old_answers": 256,
        "common_old_successes_previous": 159,
        "common_old_successes_current": 164,
        "review_questions": 12,
        "new_questions": 12,
        "effective_questions": 18,
        "all_correct_questions": 17,
        "all_wrong_questions": 9,
        "effective_new_questions": 4,
        "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
        "new_difficulty_by_kind": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 1,
            "effective_questions": 1,
            "minimum_actions_delta_range": [
              -5,
              -4
            ],
            "questions": 2,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 1,
            "all_succeeded": 0,
            "effective_questions": 3,
            "minimum_actions_delta_range": [
              2,
              4
            ],
            "questions": 4,
            "required_chain_depth_delta_range": [
              1,
              1
            ]
          },
          "wide": {
            "all_failed": 2,
            "all_succeeded": 4,
            "effective_questions": 0,
            "minimum_actions_delta_range": [
              -1,
              1
            ],
            "questions": 6,
            "required_chain_depth_delta_range": [
              -1,
              0
            ]
          }
        },
        "signal_coverage": {
          "failed_questions": 20,
          "first_answer_failed_questions": 10,
          "independent_student_answers": 0,
          "later_failures_considered": 10,
          "mixed_source_questions": 14,
          "regressed_question_ids": [
            "generated:local_textcraft_188",
            "generated:local_textcraft_248",
            "generated:local_textcraft_280",
            "generated:local_textcraft_462",
            "generated:local_textcraft_502",
            "generated:local_textcraft_517"
          ],
          "source_questions": 32,
          "upward_sources_skipped_without_teacher_proposals": 1
        },
        "main_learning_updates": 1,
        "repair_learning_updates": 1,
        "new_learning_updates_completed": 2,
        "main_learning_complete": true,
        "repair_learning_complete": true,
        "main_model_saved": true,
        "repair_model_saved": true,
        "model_saved": true,
        "main_learning_reused": false,
        "new_branch_answers": 352,
        "collection_policy_id": "64bba71af88c1427cc676e23f26a83f3f6f70ead23698dae58c93bd283d64fe7",
        "main_policy_id": "f445a3e252ec14df358c1fc37bb26fd7c0756cce8034ad607926b0642b7330a7",
        "final_policy_id": "aeca99455724c3b1855ef6b0ed30ea2fce7dd20dfd01f7a907d386b0c2667ce3",
        "previous_round_final_model_matches_collection_model": true,
        "answers_collected_before_third_learning": true,
        "completed_main_learning_rounds_before_collection": 2,
        "answer_meaning": "第三轮答卷在本轮主练和纠错前收集，反映第二轮最终模型。第三轮最终效果待共同30题测试。",
        "effect_interpretation": "固定20道种子由125/160降至118/160；共同32道旧题由159/256升至164/256，两个范围的变化方向不同。",
        "repair": {
          "accepted_by_mode": {
            "prevention": 31,
            "recovery": 15
          },
          "accepted_proposals": 46,
          "optimizer_steps": 1,
          "proposal_count": 69,
          "rejection_reasons": {
            "NO_SUCCESSFUL_RELEVANT_CORRECTION": 2,
            "SUFFIX_DID_NOT_COMPLETE_ORIGINAL_TASK": 21
          },
          "repaired_target_ids": [
            "generated:cluster_wdt_r1_s0",
            "generated:cluster_wdt_r1_s18",
            "generated:cluster_wdt_repair_r2_s12",
            "generated:cluster_wdt_repair_r2_s2",
            "generated:cluster_wdt_repair_r2_s3",
            "generated:cluster_wdt_repair_r2_s5",
            "generated:cluster_wdt_repair_r3_s21",
            "generated:cluster_wdt_repair_r3_s23",
            "generated:cluster_wdt_repair_r3_s5",
            "generated:cluster_wdt_repair_r3_s7",
            "generated:local_textcraft_188",
            "generated:local_textcraft_248",
            "generated:local_textcraft_280",
            "generated:local_textcraft_448",
            "generated:local_textcraft_462",
            "generated:local_textcraft_489",
            "generated:local_textcraft_502",
            "generated:local_textcraft_517"
          ],
          "selected_actions": 73,
          "original_target_count": 23,
          "original_goal_completed_proposals": 47,
          "accepted_target_count": 18,
          "completed_original_goal_without_accepted_actions": 1,
          "selected_correction_tokens": 1055,
          "original_target_cases": [
            {
              "task_id": "generated:local_textcraft_477",
              "label": "书架",
              "proposals": 3,
              "original_goal_completed_proposals": 0,
              "accepted_proposals": 0
            },
            {
              "task_id": "generated:local_textcraft_489",
              "label": "完整红旗",
              "proposals": 3,
              "original_goal_completed_proposals": 3,
              "accepted_proposals": 3
            },
            {
              "task_id": "generated:local_textcraft_188",
              "label": "安山岩",
              "proposals": 3,
              "original_goal_completed_proposals": 3,
              "accepted_proposals": 2
            }
          ],
          "original_goal_completion_definition": "原记录 judge_status 为 PASS；与是否存在合格学习动作分别统计。",
          "accepted_proposals_definition": "完成原目标并通过改正动作筛选的方案。",
          "selected_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
          "student_independent_success_after_learning_verified": false
        },
        "final_common30_test_completed": false,
        "complete_round_seconds": 5756.498396,
        "stage_seconds": {
          "plan-round": 2566.931565,
          "train-round": 3189.53828
        },
        "pure_parameter_training_seconds": null,
        "timing_scope": "完整一轮，含出题、作答、主练、纠错、保存和阶段切换。",
        "stage_observation_intervals": {
          "remaining_answers_and_preparation_seconds_interval": [
            696.882401,
            743.968748
          ],
          "main_learning_plus_saving_seconds_interval": [
            450.458368,
            543.847943
          ],
          "repair_material_generation_and_validation_seconds_interval": [
            1892.11893,
            1977.927737
          ],
          "repair_learning_saving_and_exit_seconds_interval": [
            17.183427,
            56.689006
          ]
        },
        "stage_observation_interval_scope": "相邻观察确定的时间范围，不是纯模型计算时间；区间上限不能直接相加。",
        "question_results": [
          {
            "task_id": "generated:local_textcraft_188",
            "section": "seed",
            "kind": "root",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_215",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_248",
            "section": "seed",
            "kind": "root",
            "successes": 3,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_273",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_280",
            "section": "seed",
            "kind": "root",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_42",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_448",
            "section": "seed",
            "kind": "root",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_462",
            "section": "seed",
            "kind": "root",
            "successes": 3,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_477",
            "section": "seed",
            "kind": "root",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_480",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_489",
            "section": "seed",
            "kind": "root",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_502",
            "section": "seed",
            "kind": "root",
            "successes": 4,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_517",
            "section": "seed",
            "kind": "root",
            "successes": 6,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:local_textcraft_328",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_357",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_100",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_72",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_45",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_190",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:local_textcraft_355",
            "section": "seed",
            "kind": "root",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_r1_s18",
            "section": "review",
            "kind": "wide",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s4",
            "section": "review",
            "kind": "wide",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_r1_s5",
            "section": "review",
            "kind": "deep_down",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s10",
            "section": "review",
            "kind": "deep_down",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s5",
            "section": "review",
            "kind": "wide",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_r1_s1",
            "section": "review",
            "kind": "deep_down",
            "successes": 4,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s17",
            "section": "review",
            "kind": "deep_down",
            "successes": 6,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s3",
            "section": "review",
            "kind": "wide",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_r1_s0",
            "section": "review",
            "kind": "wide",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s12",
            "section": "review",
            "kind": "deep_up",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s14",
            "section": "review",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r2_s2",
            "section": "review",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s0",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s1",
            "section": "new",
            "kind": "deep_down",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s16",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s17",
            "section": "new",
            "kind": "deep_down",
            "successes": 7,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s4",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s5",
            "section": "new",
            "kind": "deep_up",
            "successes": 5,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s20",
            "section": "new",
            "kind": "wide",
            "successes": 8,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s21",
            "section": "new",
            "kind": "deep_up",
            "successes": 2,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s6",
            "section": "new",
            "kind": "wide",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s7",
            "section": "new",
            "kind": "deep_up",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s22",
            "section": "new",
            "kind": "wide",
            "successes": 0,
            "attempts": 8,
            "has_signal": false,
            "included_in_update": false
          },
          {
            "task_id": "generated:cluster_wdt_repair_r3_s23",
            "section": "new",
            "kind": "deep_up",
            "successes": 3,
            "attempts": 8,
            "has_signal": true,
            "included_in_update": true
          }
        ],
        "source_records": {
          "CAPTURE.json": {
            "bytes": 58292,
            "sha256": "481c8837533874f3fcc9d252a3779276d4967bba13aaf908273090373464d3c0"
          },
          "complete.json": {
            "bytes": 9988,
            "sha256": "94889594a144983fd0a0bd53ce66a79f38c72aeb7b8218d4be6868492535af1e"
          },
          "collection.json": {
            "bytes": 12803,
            "sha256": "05ccda0149e1ca38cabe05bb2db12793d5cda60a5853ef2ecaa99a684e720b6f"
          },
          "plan.json": {
            "bytes": 101855,
            "sha256": "cf4a8aac703816624a1ad4bc09cc51656228a6ff920b5947ab9f1286b220e0c8"
          },
          "main.json": {
            "bytes": 8330,
            "sha256": "80ea9fbc1874462bb70cd08e5b134e9ead463148db9d672c67b66e6020b6a759"
          },
          "repair.json": {
            "bytes": 952,
            "sha256": "a9aac9f467d3dd314e804ea9a0205150d5607d8ca7928fd4b9ec097e4a6008c5"
          },
          "repair_summary.json": {
            "bytes": 116040,
            "sha256": "3c50f288bb48e2ec1440abfa4794cfc601f79c6ba574a786780a675d4dbab041"
          },
          "main_model/COMPLETE.json": {
            "bytes": 337,
            "sha256": "9af78ea8fced779502f596ff16b3076f12dc05de0324b1110ac2a350e4649d99"
          },
          "repair_model/COMPLETE.json": {
            "bytes": 337,
            "sha256": "938db3362a51c2379bf21664056a7281fe2cfe67a92f2a47b8dad82bd397ea49"
          },
          "timing/WDT_REPAIR_plan-round_3_attempt_01.started.json": {
            "bytes": 1141,
            "sha256": "b9c492335e3fbf15cb8a3a29143d5b1d0416666fd571fce252f7f02e72208b5c"
          },
          "timing/WDT_REPAIR_plan-round_3_attempt_01.exit.json": {
            "bytes": 180,
            "sha256": "5dc4349ad5667c350da7433bf1d232e41b7b24498a636f6de6f52b4c60d5d19f"
          },
          "timing/WDT_REPAIR_train-round_3_attempt_01.started.json": {
            "bytes": 1148,
            "sha256": "76bf2a00a9f18084a8d6b25eb69acf9225e24c4cf8e08aaaffe85c9dc412edc5"
          },
          "timing/WDT_REPAIR_train-round_3_attempt_01.exit.json": {
            "bytes": 181,
            "sha256": "e6fe2939a08c4705ad4a143a9e4599d52f6d6b8a92c86fae887284913c6153e1"
          }
        },
        "parent_analysis": {
          "bytes": 8575,
          "sha256": "559e0778b28fff5ece438ae2b57f577dfd0d50c06e2ffe4829c427fd68a9e80b"
        }
      },
      "training_complete": true
    },
    {
      "label": "WDT_SCALE50",
      "display_label": "SCALE50",
      "seed_questions": 50,
      "round_question_caps": [
        100,
        100,
        100
      ],
      "target_questions": null,
      "first_main_round_source": "initial_untrained_model",
      "new_main_rounds": [
        1,
        2,
        3
      ],
      "new_main_answers_upper_bound": 2000,
      "main_updates_per_round": 1,
      "repair": false,
      "reviews_per_round_max": 25,
      "seed_selection": "retain original20 plus fixed-random30 from prepared official training pool; no student or test scores",
      "training_started": true,
      "status": "TRAINING_COMPLETE",
      "status_label": "三轮已完成并保存",
      "current_phase": "waiting_final_test",
      "new_learning_started": true,
      "new_learning_updates_completed": 3,
      "imported_main_learning": false,
      "reused_first_round_questions": 0,
      "reused_first_round_answers": 0,
      "new_main_answers_saved": 1872,
      "round_progress": [
        {
          "round": 1,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 1,
          "questions": 75,
          "answers_saved": 600,
          "effective_questions": 37,
          "model_saved": true
        },
        {
          "round": 2,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 1,
          "questions": 75,
          "answers_saved": 600,
          "effective_questions": 40,
          "model_saved": true
        },
        {
          "round": 3,
          "status": "COMPLETE",
          "imported_main_learning": false,
          "new_learning_updates_completed": 1,
          "questions": 84,
          "answers_saved": 672,
          "effective_questions": 40,
          "model_saved": true,
          "question_cap": 100,
          "complete_question_groups": 84,
          "current_phase": "complete"
        }
      ],
      "current_answers_required": 0,
      "current_answers_upper_bound": 0,
      "initial_three_round_answer_ceiling": 2400,
      "remaining_new_main_answers_upper_bound": 0,
      "observed_round_questions": [
        75,
        75,
        84
      ],
      "first_round_result": {
        "arm": "WDT_SCALE50",
        "round": 1,
        "status": "COMPLETE",
        "completed_at_utc": "2026-11-02T00:05:42.738445+00:00",
        "completed_at_beijing": "2026-11-02 08:05:42",
        "questions": 75,
        "answers": 600,
        "answers_per_question": 8,
        "seed_questions": 50,
        "seed_answers": 400,
        "review_questions": 0,
        "new_questions": 25,
        "new_answers": 200,
        "effective_questions": 37,
        "effective_questions_definition": "八份回答有对有错，实际参与本轮学习的题数。",
        "effective_new_questions": 13,
        "all_correct_new_questions": 9,
        "all_wrong_new_questions": 3,
        "new_difficulty_by_kind": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 5,
            "effective_questions": 3,
            "minimum_actions_delta_range": [
              -18,
              -3
            ],
            "questions": 8,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 1,
            "all_succeeded": 2,
            "effective_questions": 6,
            "minimum_actions_delta_range": [
              1,
              6
            ],
            "questions": 9,
            "required_chain_depth_delta_range": [
              1,
              2
            ]
          },
          "wide": {
            "all_failed": 2,
            "all_succeeded": 2,
            "effective_questions": 4,
            "minimum_actions_delta_range": [
              -1,
              1
            ],
            "questions": 8,
            "required_chain_depth_delta_range": [
              -1,
              1
            ]
          }
        },
        "signal_coverage": {
          "failed_questions": 30,
          "first_answer_failed_questions": 16,
          "independent_student_answers": 0,
          "later_failures_considered": 14,
          "mixed_source_questions": 24,
          "regressed_question_ids": [],
          "source_questions": 50,
          "upward_sources_skipped_without_teacher_proposals": 2
        },
        "answers_model": "initial_untrained_model",
        "answers_collected_before_first_learning": true,
        "completed_learning_rounds_before_collection": 0,
        "main_learning_complete": true,
        "model_saved": true,
        "new_learning_updates_completed": 1,
        "final_policy_id": "fab4264a5e01094a423b60df09b4fa894608e2e7c3be889572cd5eeb57784da0",
        "timing": {
          "complete_round_seconds": 6320.476063,
          "planning_and_seed_answers_seconds": 3982.668088,
          "new_answers_and_setup_seconds_interval": [
            1260.311598,
            1305.488575
          ],
          "learning_and_save_seconds_interval": [
            1032.292323,
            1077.4693
          ]
        },
        "result_meaning": "这批答卷由初始模型完成，用于说明首轮练习是否提供成败比较；训练后的提升要看后续作答和共同测试。"
      },
      "second_round_result": {
        "arm": "WDT_SCALE50",
        "round": 2,
        "status": "COMPLETE",
        "questions": 75,
        "answers": 600,
        "seed_questions": 50,
        "seed_answers": 400,
        "seed_successes": 279,
        "review_questions": 25,
        "new_questions": 0,
        "effective_questions": 40,
        "final_policy_id": "20c9a42e5b230319dadf41a42c85c32b7c148dead5b0a32fe9a499c0b098070f",
        "completed_at_utc": "2026-11-02T03:44:15.742210+00:00",
        "completed_at_beijing": "2026-11-02 11:44:15",
        "started_at_utc": "2026-11-02T02:13:21.078609+00:00",
        "complete_round_seconds": 5454.663601,
        "stage_seconds": {
          "plan-round": 4271.307887,
          "train-round": 1183.324984
        },
        "pure_parameter_training_seconds": null,
        "all_correct_questions": 28,
        "all_wrong_questions": 7,
        "effective_new_questions": 0,
        "new_difficulty_by_kind": {},
        "signal_coverage": {
          "failed_questions": 47,
          "first_answer_failed_questions": 27,
          "independent_student_answers": 0,
          "later_failures_considered": 20,
          "mixed_source_questions": 40,
          "regressed_question_ids": [
            "generated:local_textcraft_248",
            "generated:local_textcraft_280",
            "generated:local_textcraft_448",
            "generated:local_textcraft_357",
            "generated:local_textcraft_190",
            "generated:random_r3_textcraft_219",
            "generated:matched_random_r3_textcraft_282",
            "generated:matched_random_r2_textcraft_304",
            "generated:matched_random_r3_textcraft_350",
            "generated:matched_random_r2_textcraft_401",
            "generated:cluster_wdt_scale50_r1_s12",
            "generated:cluster_wdt_scale50_r1_s11",
            "generated:cluster_wdt_scale50_r1_s10",
            "generated:cluster_wdt_scale50_r1_s20",
            "generated:cluster_wdt_scale50_r1_s35",
            "generated:cluster_wdt_scale50_r1_s1"
          ],
          "source_questions": 75,
          "upward_sources_skipped_without_teacher_proposals": 0
        },
        "reserved_new_question_positions": 25,
        "main_learning_updates": 1,
        "main_learning_complete": true,
        "model_saved": true,
        "repair": null,
        "answers_collected_before_second_learning": true,
        "completed_main_learning_rounds_before_collection": 1,
        "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
        "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
        "source_records": {
          "CAPTURE.json": {
            "bytes": 81497,
            "sha256": "91e067ae2740e14e5df137b4df9fbc001f52428b422b65b7df3d1c5be57bd069"
          },
          "complete.json": {
            "bytes": 13748,
            "sha256": "3a3719b6addf1e63cdbfbcbeea0f32e15ee74e41c34011bb198dc13ca9e63ad5"
          },
          "collection.json": {
            "bytes": 21637,
            "sha256": "9d7fc5d35cc7a9980483f1231b04b63ada06f4a01d5b47c916fbac52e6ee613a"
          },
          "plan.json": {
            "bytes": 126053,
            "sha256": "6b3cc77887a2b7503bad20e03b4a6242b8a30475a7e17a641b270d26d2613d34"
          },
          "main.json": {
            "bytes": 13037,
            "sha256": "d94176b7a4f9d90b43afb37b46d8e6394f2ae7192519fd6961ebe70b4e8dcb1f"
          }
        },
        "new_question_issue": {
          "valid_individual_diagnoses": 55,
          "partition_assigned_diagnoses": 54,
          "omitted_diagnosis_id": "D44",
          "generation_candidate_requests": 0,
          "reason": "老师分组漏掉一条诊断，整份分组被拒，55 份有效单题诊断未进入出题。",
          "training_continued_on_existing_questions": true,
          "successful_expansion_this_round": false,
          "evidence": {
            "bytes": 1367166,
            "sha256": "b710a6d99f7f90af62c0fe62af62e5529354b0c5456e415d7e79660b9ef59f39"
          }
        }
      },
      "active_round": null,
      "current_completed_question_groups": 84,
      "completed_new_main_answers_saved": 1872,
      "current_new_main_answers_saved": 0,
      "current_answer_count_available_in_snapshot": true,
      "training_complete": true,
      "actual_questions_per_round": [
        75,
        75,
        84
      ],
      "third_round_result": {
        "arm": "WDT_SCALE50",
        "round": 3,
        "status": "COMPLETE",
        "seed_questions": 50,
        "review_questions": 25,
        "new_questions": 9,
        "questions": 84,
        "answers": 672,
        "answers_per_question": 8,
        "effective_questions": 40,
        "effective_new_questions": 4,
        "all_succeeded_questions": 33,
        "all_failed_questions": 11,
        "new_main_learning_updates": 1,
        "final_model_saved": true,
        "repair_enabled": false,
        "new_question_breakdown": {
          "deep_down": {
            "all_failed": 0,
            "all_succeeded": 1,
            "effective_questions": 0,
            "minimum_actions_delta_range": [
              -10,
              -10
            ],
            "questions": 1,
            "required_chain_depth_delta_range": [
              -1,
              -1
            ]
          },
          "deep_up": {
            "all_failed": 0,
            "all_succeeded": 0,
            "effective_questions": 2,
            "minimum_actions_delta_range": [
              1,
              4
            ],
            "questions": 2,
            "required_chain_depth_delta_range": [
              1,
              1
            ]
          },
          "wide": {
            "all_failed": 1,
            "all_succeeded": 3,
            "effective_questions": 2,
            "minimum_actions_delta_range": [
              -3,
              1
            ],
            "questions": 6,
            "required_chain_depth_delta_range": [
              -1,
              1
            ]
          }
        },
        "seed_observations": {
          "questions": 50,
          "previous_round": 2,
          "current_round": 3,
          "previous_successes": 279,
          "previous_answers": 400,
          "current_successes": 279,
          "current_answers": 400,
          "improved_questions": 12,
          "regressed_questions": 17,
          "unchanged_questions": 21,
          "recovered_from_all_failed": 0,
          "regressed_to_all_failed": 2
        },
        "common_old_task_observations": {
          "questions": 75,
          "previous_round": 2,
          "current_round": 3,
          "previous_successes": 402,
          "previous_answers": 600,
          "current_successes": 407,
          "current_answers": 600,
          "improved_questions": 20,
          "regressed_questions": 22,
          "unchanged_questions": 33,
          "recovered_from_all_failed": 1,
          "regressed_to_all_failed": 4
        },
        "observation_interpretation": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
        "start_utc": "2026-11-02T06:22:35.695521+00:00",
        "end_utc": "2026-11-02T08:10:21.268705+00:00",
        "start_beijing": "2026-11-02 14:22:35",
        "end_beijing": "2026-11-02 16:10:21",
        "full_elapsed_seconds": 6465.573184,
        "stage_sum_seconds": 6465.546611000001,
        "interstage_seconds": 0.026573,
        "recorded_stage_seconds": {
          "plan-round": 4874.832532,
          "train-round": 1590.714079
        },
        "pure_parameter_training_seconds": null,
        "elapsed_scope": "整轮用时包含出题、作答、学习、保存及环节交接；后半段也不是纯模型调整时间。",
        "schema_version": "wide_deep_SCALE50_round3_summary_v1",
        "completed_at_utc": "2026-11-02T08:10:21.268705+00:00",
        "completed_at_beijing": "2026-11-02 16:10:21",
        "started_at_utc": "2026-11-02T06:22:35.695521+00:00",
        "complete_round_seconds": 6465.573184,
        "model_saved": true,
        "main_learning_complete": true,
        "main_learning_updates": 1,
        "new_learning_updates_completed": 1,
        "answers_collected_before_third_learning": true,
        "completed_main_learning_rounds_before_collection": 2,
        "answer_meaning": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
        "final_common30_test_completed": false,
        "source_public_summary": "scale50-round-3-summary.json"
      }
    }
  ],
  "third_round_common_seed_comparison": {
    "questions": 20,
    "answers_per_arm": 160,
    "successes": {
      "WT": 121,
      "DT": 121,
      "WDT": 120,
      "R": 120
    },
    "completed_learning_rounds_before_collection": 2,
    "collected_answers_precede_third_round_learning": true,
    "nonseed_question_sets_differ": true,
    "full_batch_totals_comparable": false,
    "interpretation": "共同种子上四组答对份数接近；其余训练题不同，全批答对总数不能用于组间排名。"
  },
  "current_answers_upper_bound": 90,
  "REPAIR_first_round_correction": {
    "arm": "WDT_REPAIR",
    "round": 1,
    "phase": "correction",
    "status": "COMPLETE",
    "imported_main_learning": true,
    "reused_main_questions": 31,
    "reused_main_answers": 248,
    "new_learning_updates_completed": 1,
    "updated": true,
    "model_saved": true,
    "proposal_count": 39,
    "accepted_proposals": 34,
    "repaired_original_targets": 12,
    "selected_learning_actions": 49,
    "selected_correction_tokens": 710,
    "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
    "original_target_cases": [
      {
        "task_id": "generated:local_textcraft_489",
        "goal": "minecraft:red_banner",
        "proposals": 3,
        "accepted_proposals": 3,
        "selected_learning_actions": 4
      },
      {
        "task_id": "generated:local_textcraft_477",
        "goal": "minecraft:bookshelf",
        "proposals": 3,
        "accepted_proposals": 0,
        "selected_learning_actions": 0
      }
    ],
    "stage_seconds": 864.250016,
    "stage_time_scope": "纠错收集、学习、保存和准备在内的完整首轮阶段。",
    "student_independent_success_after_learning_verified": false,
    "final_policy_id": "5ad18b4b0b35fb42026592d3650659ad39bdc6e65efd8d03a03bab7dfe1e5d14"
  },
  "observation_history": [
    {
      "status": {
        "label": "REPAIR 首轮纠错已完成；SCALE50 首轮出题与作答中",
        "page_is_a_live_monitor": false,
        "observed_at_utc": "2026-11-01T22:22:14.613781+00:00",
        "observed_at_beijing": "2026-11-02 06:22:14",
        "original_formal_training_started": true,
        "original_active_revision": "teacher_cluster_deep_history_v6",
        "active_arm": "WDT_SCALE50",
        "active_round": 1,
        "active_phase": "generation",
        "active_answers_saved": 5,
        "active_answers_required": null,
        "completed_learning_rounds": 12,
        "original_required_learning_rounds": 12,
        "completed_original_full_training_arms": [
          "WT",
          "DT",
          "WDT",
          "R"
        ],
        "completed_experiment_rounds": 3,
        "extension_release_ready": true,
        "extension_deployed": true,
        "extension_training_started": true,
        "new_method_results_available": false,
        "final_evaluation_started": false,
        "final_evaluation_completed": false,
        "billing_stopped_confirmed": false,
        "reported_monitor_error": null,
        "extension_prepared_and_attached": true,
        "extension_controller_started": true,
        "extension_waits_for_original_round3_checkpoints": false,
        "active_answer_count_available_in_snapshot": true,
        "active_revision": "curriculum_repair_release_v7",
        "active_phase_label": "首轮出题与作答中",
        "active_answers_origin": "new_SCALE50_round_1",
        "active_new_main_answers_saved": 5,
        "active_imported_main_learning": false,
        "completed_learning_rounds_scope": "original_four_groups",
        "original_training_complete": true,
        "extension_started_at_utc": "2026-11-01T22:05:55.929321+00:00",
        "extension_started_at_beijing": "2026-11-02 06:05:55",
        "training_started_definition": "本组训练流程已经开始；复用首轮记录不计为新增学习。",
        "extension_new_learning_started": true,
        "new_learning_updates_completed": 1,
        "main_test_answers_saved": 0,
        "auxiliary_test_answers_saved": 0,
        "active_answers_upper_bound": 800,
        "active_question_cap": 100,
        "extension_learning_results_available": true,
        "new_method_results_scope": "independent_comparison_results",
        "latest_completed_repair": {
          "arm": "WDT_REPAIR",
          "round": 1,
          "phase": "correction",
          "status": "COMPLETE",
          "imported_main_learning": true,
          "reused_main_questions": 31,
          "reused_main_answers": 248,
          "new_learning_updates_completed": 1,
          "updated": true,
          "model_saved": true,
          "proposal_count": 39,
          "accepted_proposals": 34,
          "repaired_original_targets": 12,
          "selected_learning_actions": 49,
          "selected_correction_tokens": 710,
          "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
          "original_target_cases": [
            {
              "task_id": "generated:local_textcraft_489",
              "goal": "minecraft:red_banner",
              "proposals": 3,
              "accepted_proposals": 3,
              "selected_learning_actions": 4
            },
            {
              "task_id": "generated:local_textcraft_477",
              "goal": "minecraft:bookshelf",
              "proposals": 3,
              "accepted_proposals": 0,
              "selected_learning_actions": 0
            }
          ],
          "stage_seconds": 864.250016,
          "stage_time_scope": "纠错收集、学习、保存和准备在内的完整首轮阶段。",
          "student_independent_success_after_learning_verified": false,
          "final_policy_id": "5ad18b4b0b35fb42026592d3650659ad39bdc6e65efd8d03a03bab7dfe1e5d14"
        }
      },
      "extension_progress": [
        {
          "label": "WDT_OPT",
          "status": "IMPORTED_FIRST_ROUND_WAITING_ROUND_2",
          "status_label": "首轮复用已接入，等待第二轮",
          "current_phase": "waiting_round_2",
          "training_started": false,
          "imported_main_learning": true,
          "reused_first_round_answers": 248,
          "new_main_answers_saved": 0,
          "new_learning_started": false,
          "new_learning_updates_completed": 0,
          "round_progress": [
            {
              "round": 1,
              "status": "COMPLETE",
              "imported_main_learning": true,
              "new_learning_updates_completed": 0
            },
            {
              "round": 2,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            },
            {
              "round": 3,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            }
          ]
        },
        {
          "label": "WDT_REPAIR",
          "status": "FIRST_ROUND_COMPLETE_WAITING_ROUND_2",
          "status_label": "首轮纠错已完成，等待第二轮",
          "current_phase": "waiting_round_2",
          "training_started": true,
          "imported_main_learning": true,
          "reused_first_round_answers": 248,
          "new_main_answers_saved": 0,
          "new_learning_started": true,
          "new_learning_updates_completed": 1,
          "round_progress": [
            {
              "round": 1,
              "status": "COMPLETE",
              "imported_main_learning": true,
              "new_learning_updates_completed": 1
            },
            {
              "round": 2,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            },
            {
              "round": 3,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            }
          ]
        },
        {
          "label": "WDT_SCALE50",
          "status": "FIRST_ROUND_GENERATION_IN_PROGRESS",
          "status_label": "首轮出题与作答中",
          "current_phase": "generation",
          "training_started": true,
          "imported_main_learning": false,
          "reused_first_round_answers": 0,
          "new_main_answers_saved": 5,
          "new_learning_started": false,
          "new_learning_updates_completed": 0,
          "round_progress": [
            {
              "round": 1,
              "status": "ACTIVE",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            },
            {
              "round": 2,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            },
            {
              "round": 3,
              "status": "PENDING",
              "imported_main_learning": false,
              "new_learning_updates_completed": 0
            }
          ]
        }
      ],
      "completed_new_learning_stages": [
        {
          "arm": "WDT_REPAIR",
          "round": 1,
          "phase": "correction",
          "status": "COMPLETE",
          "imported_main_learning": true,
          "reused_main_questions": 31,
          "reused_main_answers": 248,
          "new_learning_updates_completed": 1,
          "updated": true,
          "model_saved": true,
          "proposal_count": 39,
          "accepted_proposals": 34,
          "repaired_original_targets": 12,
          "selected_learning_actions": 49,
          "selected_correction_tokens": 710,
          "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
          "original_target_cases": [
            {
              "task_id": "generated:local_textcraft_489",
              "goal": "minecraft:red_banner",
              "proposals": 3,
              "accepted_proposals": 3,
              "selected_learning_actions": 4
            },
            {
              "task_id": "generated:local_textcraft_477",
              "goal": "minecraft:bookshelf",
              "proposals": 3,
              "accepted_proposals": 0,
              "selected_learning_actions": 0
            }
          ],
          "stage_seconds": 864.250016,
          "stage_time_scope": "纠错收集、学习、保存和准备在内的完整首轮阶段。",
          "student_independent_success_after_learning_verified": false,
          "final_policy_id": "5ad18b4b0b35fb42026592d3650659ad39bdc6e65efd8d03a03bab7dfe1e5d14"
        }
      ]
    },
    {
      "status": {
        "label": "SCALE50 首轮已完成；OPT 第二轮出题与作答中",
        "page_is_a_live_monitor": false,
        "observed_at_utc": "2026-11-02T00:08:00.598508+00:00",
        "observed_at_beijing": "2026-11-02 08:08:00",
        "original_formal_training_started": true,
        "original_active_revision": "teacher_cluster_deep_history_v6",
        "active_arm": "WDT_OPT",
        "active_round": 2,
        "active_phase": "generation",
        "active_answers_saved": 10,
        "active_answers_required": 320,
        "completed_learning_rounds": 12,
        "original_required_learning_rounds": 12,
        "completed_original_full_training_arms": [
          "WT",
          "DT",
          "WDT",
          "R"
        ],
        "completed_experiment_rounds": 3,
        "extension_release_ready": true,
        "extension_deployed": true,
        "extension_training_started": true,
        "new_method_results_available": false,
        "final_evaluation_started": false,
        "final_evaluation_completed": false,
        "billing_stopped_confirmed": false,
        "reported_monitor_error": null,
        "extension_prepared_and_attached": true,
        "extension_controller_started": true,
        "extension_waits_for_original_round3_checkpoints": false,
        "active_answer_count_available_in_snapshot": true,
        "active_revision": "curriculum_repair_release_v7",
        "active_phase_label": "第二轮出题与作答中",
        "active_answers_origin": "new_OPT_round_2",
        "active_new_main_answers_saved": 10,
        "active_imported_main_learning": false,
        "completed_learning_rounds_scope": "original_four_groups",
        "original_training_complete": true,
        "extension_started_at_utc": "2026-11-01T22:05:55.929321+00:00",
        "extension_started_at_beijing": "2026-11-02 06:05:55",
        "training_started_definition": "本组训练流程已经开始；复用首轮记录不计为新增学习。",
        "extension_new_learning_started": true,
        "new_learning_updates_completed": 2,
        "main_test_answers_saved": 0,
        "auxiliary_test_answers_saved": 0,
        "active_answers_upper_bound": 320,
        "active_question_cap": 60,
        "extension_learning_results_available": true,
        "new_method_results_scope": "independent_comparison_results",
        "latest_completed_repair": {
          "arm": "WDT_REPAIR",
          "round": 1,
          "phase": "correction",
          "status": "COMPLETE",
          "imported_main_learning": true,
          "reused_main_questions": 31,
          "reused_main_answers": 248,
          "new_learning_updates_completed": 1,
          "updated": true,
          "model_saved": true,
          "proposal_count": 39,
          "accepted_proposals": 34,
          "repaired_original_targets": 12,
          "selected_learning_actions": 49,
          "selected_correction_tokens": 710,
          "selected_learning_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
          "original_target_cases": [
            {
              "task_id": "generated:local_textcraft_489",
              "goal": "minecraft:red_banner",
              "proposals": 3,
              "accepted_proposals": 3,
              "selected_learning_actions": 4
            },
            {
              "task_id": "generated:local_textcraft_477",
              "goal": "minecraft:bookshelf",
              "proposals": 3,
              "accepted_proposals": 0,
              "selected_learning_actions": 0
            }
          ],
          "stage_seconds": 864.250016,
          "stage_time_scope": "纠错收集、学习、保存和准备在内的完整首轮阶段。",
          "student_independent_success_after_learning_verified": false,
          "final_policy_id": "5ad18b4b0b35fb42026592d3650659ad39bdc6e65efd8d03a03bab7dfe1e5d14"
        },
        "latest_completed_scale50": {
          "arm": "WDT_SCALE50",
          "round": 1,
          "status": "COMPLETE",
          "completed_at_utc": "2026-11-02T00:05:42.738445+00:00",
          "completed_at_beijing": "2026-11-02 08:05:42",
          "questions": 75,
          "answers": 600,
          "answers_per_question": 8,
          "seed_questions": 50,
          "seed_answers": 400,
          "review_questions": 0,
          "new_questions": 25,
          "new_answers": 200,
          "effective_questions": 37,
          "effective_questions_definition": "八份回答有对有错，实际参与本轮学习的题数。",
          "effective_new_questions": 13,
          "all_correct_new_questions": 9,
          "all_wrong_new_questions": 3,
          "new_difficulty_by_kind": {
            "deep_down": {
              "all_failed": 0,
              "all_succeeded": 5,
              "effective_questions": 3,
              "minimum_actions_delta_range": [
                -18,
                -3
              ],
              "questions": 8,
              "required_chain_depth_delta_range": [
                -1,
                -1
              ]
            },
            "deep_up": {
              "all_failed": 1,
              "all_succeeded": 2,
              "effective_questions": 6,
              "minimum_actions_delta_range": [
                1,
                6
              ],
              "questions": 9,
              "required_chain_depth_delta_range": [
                1,
                2
              ]
            },
            "wide": {
              "all_failed": 2,
              "all_succeeded": 2,
              "effective_questions": 4,
              "minimum_actions_delta_range": [
                -1,
                1
              ],
              "questions": 8,
              "required_chain_depth_delta_range": [
                -1,
                1
              ]
            }
          },
          "signal_coverage": {
            "failed_questions": 30,
            "first_answer_failed_questions": 16,
            "independent_student_answers": 0,
            "later_failures_considered": 14,
            "mixed_source_questions": 24,
            "regressed_question_ids": [],
            "source_questions": 50,
            "upward_sources_skipped_without_teacher_proposals": 2
          },
          "answers_model": "initial_untrained_model",
          "answers_collected_before_first_learning": true,
          "completed_learning_rounds_before_collection": 0,
          "main_learning_complete": true,
          "model_saved": true,
          "new_learning_updates_completed": 1,
          "final_policy_id": "fab4264a5e01094a423b60df09b4fa894608e2e7c3be889572cd5eeb57784da0",
          "timing": {
            "complete_round_seconds": 6320.476063,
            "planning_and_seed_answers_seconds": 3982.668088,
            "new_answers_and_setup_seconds_interval": [
              1260.311598,
              1305.488575
            ],
            "learning_and_save_seconds_interval": [
              1032.292323,
              1077.4693
            ]
          },
          "result_meaning": "这批答卷由初始模型完成，用于说明首轮练习是否提供成败比较；训练后的提升要看后续作答和共同测试。"
        }
      }
    },
    {
      "observed_at_utc": "2026-11-02T04:20:03.369810+00:00",
      "observed_at_beijing": "2026-11-02 12:20:03",
      "completed_all_training_rounds": 18,
      "active_arm": "WDT_OPT",
      "active_round": 3,
      "current_answers_saved": 256,
      "current_answers_required": 352
    },
    {
      "observed_at_utc": "2026-11-02T04:48:07.580166+00:00",
      "observed_at_beijing": "2026-11-02 12:48:07",
      "completed_all_training_rounds": 19,
      "active_arm": "WDT_REPAIR",
      "active_round": 3,
      "current_answers_saved": null,
      "current_answers_required": 352
    },
    {
      "observed_at_utc": "2026-11-02T06:28:20.285682+00:00",
      "observed_at_beijing": "2026-11-02 14:28:20",
      "completed_all_training_rounds": 20,
      "active_arm": "WDT_SCALE50",
      "active_round": 3,
      "current_answers_saved": 41,
      "current_answers_required": null
    },
    {
      "observed_at_utc": "2026-11-02T08:11+00:00",
      "observed_at_beijing": "2026-11-02 16:11",
      "completed_all_training_rounds": 21,
      "active_arm": "N",
      "current_answers_saved": 0,
      "main_test_answers_saved": 0
    },
    {
      "observed_at_beijing": "2026-11-02 16:44:52",
      "active_arm": "WT",
      "current_answers_saved": 45,
      "main_test_answers_saved": 135
    },
    {
      "observed_at_beijing": "2026-11-02 17:14",
      "active_arm": "DT",
      "current_answers_saved": 65,
      "main_test_answers_saved": 239,
      "source": "wt-test-interruption-1714.json"
    }
  ],
  "SCALE50_first_round": {
    "arm": "WDT_SCALE50",
    "round": 1,
    "status": "COMPLETE",
    "completed_at_utc": "2026-11-02T00:05:42.738445+00:00",
    "completed_at_beijing": "2026-11-02 08:05:42",
    "questions": 75,
    "answers": 600,
    "answers_per_question": 8,
    "seed_questions": 50,
    "seed_answers": 400,
    "review_questions": 0,
    "new_questions": 25,
    "new_answers": 200,
    "effective_questions": 37,
    "effective_questions_definition": "八份回答有对有错，实际参与本轮学习的题数。",
    "effective_new_questions": 13,
    "all_correct_new_questions": 9,
    "all_wrong_new_questions": 3,
    "new_difficulty_by_kind": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 5,
        "effective_questions": 3,
        "minimum_actions_delta_range": [
          -18,
          -3
        ],
        "questions": 8,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 1,
        "all_succeeded": 2,
        "effective_questions": 6,
        "minimum_actions_delta_range": [
          1,
          6
        ],
        "questions": 9,
        "required_chain_depth_delta_range": [
          1,
          2
        ]
      },
      "wide": {
        "all_failed": 2,
        "all_succeeded": 2,
        "effective_questions": 4,
        "minimum_actions_delta_range": [
          -1,
          1
        ],
        "questions": 8,
        "required_chain_depth_delta_range": [
          -1,
          1
        ]
      }
    },
    "signal_coverage": {
      "failed_questions": 30,
      "first_answer_failed_questions": 16,
      "independent_student_answers": 0,
      "later_failures_considered": 14,
      "mixed_source_questions": 24,
      "regressed_question_ids": [],
      "source_questions": 50,
      "upward_sources_skipped_without_teacher_proposals": 2
    },
    "answers_model": "initial_untrained_model",
    "answers_collected_before_first_learning": true,
    "completed_learning_rounds_before_collection": 0,
    "main_learning_complete": true,
    "model_saved": true,
    "new_learning_updates_completed": 1,
    "final_policy_id": "fab4264a5e01094a423b60df09b4fa894608e2e7c3be889572cd5eeb57784da0",
    "timing": {
      "complete_round_seconds": 6320.476063,
      "planning_and_seed_answers_seconds": 3982.668088,
      "new_answers_and_setup_seconds_interval": [
        1260.311598,
        1305.488575
      ],
      "learning_and_save_seconds_interval": [
        1032.292323,
        1077.4693
      ]
    },
    "result_meaning": "这批答卷由初始模型完成，用于说明首轮练习是否提供成败比较；训练后的提升要看后续作答和共同测试。"
  },
  "completed_all_training_rounds": 21,
  "required_all_training_rounds": 21,
  "completed_extension_new_main_answers": 3216,
  "OPT_second_round": {
    "arm": "WDT_OPT",
    "round": 2,
    "status": "COMPLETE",
    "questions": 40,
    "answers": 320,
    "seed_questions": 20,
    "seed_answers": 160,
    "seed_successes": 125,
    "review_questions": 10,
    "new_questions": 10,
    "effective_questions": 17,
    "final_policy_id": "69cec642ec57251107eebc71b4920feb9db9e5c0e8a660fe9c30f070019e92e9",
    "completed_at_utc": "2026-11-02T00:59:16.661732+00:00",
    "completed_at_beijing": "2026-11-02 08:59:16",
    "started_at_utc": "2026-11-02T00:05:42.761413+00:00",
    "complete_round_seconds": 3213.900319,
    "stage_seconds": {
      "plan-round": 2116.67438,
      "train-round": 1097.208328
    },
    "pure_parameter_training_seconds": null,
    "all_correct_questions": 16,
    "all_wrong_questions": 7,
    "effective_new_questions": 4,
    "new_difficulty_by_kind": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 0,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          -17,
          -17
        ],
        "questions": 1,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 3,
        "all_succeeded": 0,
        "effective_questions": 2,
        "minimum_actions_delta_range": [
          1,
          8
        ],
        "questions": 5,
        "required_chain_depth_delta_range": [
          1,
          2
        ]
      },
      "wide": {
        "all_failed": 1,
        "all_succeeded": 2,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          0,
          1
        ],
        "questions": 4,
        "required_chain_depth_delta_range": [
          0,
          1
        ]
      }
    },
    "signal_coverage": {
      "failed_questions": 16,
      "first_answer_failed_questions": 10,
      "independent_student_answers": 0,
      "later_failures_considered": 6,
      "mixed_source_questions": 13,
      "regressed_question_ids": [
        "generated:local_textcraft_357",
        "generated:local_textcraft_190",
        "generated:cluster_wdt_r1_s4",
        "generated:cluster_wdt_r1_s18"
      ],
      "source_questions": 30,
      "upward_sources_skipped_without_teacher_proposals": 1
    },
    "reserved_new_question_positions": 10,
    "main_learning_updates": 1,
    "main_learning_complete": true,
    "model_saved": true,
    "repair": null,
    "answers_collected_before_second_learning": true,
    "completed_main_learning_rounds_before_collection": 1,
    "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
    "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
    "source_records": {
      "CAPTURE.json": {
        "bytes": 41850,
        "sha256": "a26b94fde3da906b1ea34543dda655657c43d2664c1479394c0f334e692336f5"
      },
      "complete.json": {
        "bytes": 8406,
        "sha256": "6890c386929017b6f72448edd03bd29afe02aa2f451945d58450a79aaa82dc0e"
      },
      "collection.json": {
        "bytes": 11663,
        "sha256": "21b5604ffc73ff9cb319affda280e0baf6ef9b65710b7f98abe4db37f17671a2"
      },
      "plan.json": {
        "bytes": 86396,
        "sha256": "e34be224d2fad9f691ce6ae821d1e5952031e7072705e45620d2f7fd2fb20002"
      },
      "main.json": {
        "bytes": 7702,
        "sha256": "3f6a9c3729731485b372e375f74a7d13464366c0e11400d6e4fcf909a2005147"
      }
    }
  },
  "REPAIR_second_round": {
    "arm": "WDT_REPAIR",
    "round": 2,
    "status": "COMPLETE",
    "questions": 40,
    "answers": 320,
    "seed_questions": 20,
    "seed_answers": 160,
    "seed_successes": 125,
    "review_questions": 10,
    "new_questions": 10,
    "effective_questions": 15,
    "final_policy_id": "64bba71af88c1427cc676e23f26a83f3f6f70ead23698dae58c93bd283d64fe7",
    "completed_at_utc": "2026-11-02T02:13:21.055887+00:00",
    "completed_at_beijing": "2026-11-02 10:13:21",
    "started_at_utc": "2026-11-02T00:59:16.683471+00:00",
    "complete_round_seconds": 4444.372416,
    "stage_seconds": {
      "plan-round": 1982.335677,
      "train-round": 2462.02242
    },
    "pure_parameter_training_seconds": null,
    "all_correct_questions": 17,
    "all_wrong_questions": 8,
    "effective_new_questions": 3,
    "new_difficulty_by_kind": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 1,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          -17,
          -17
        ],
        "questions": 2,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 3,
        "all_succeeded": 0,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          1,
          4
        ],
        "questions": 4,
        "required_chain_depth_delta_range": [
          1,
          1
        ]
      },
      "wide": {
        "all_failed": 1,
        "all_succeeded": 2,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          0,
          1
        ],
        "questions": 4,
        "required_chain_depth_delta_range": [
          0,
          0
        ]
      }
    },
    "signal_coverage": {
      "failed_questions": 16,
      "first_answer_failed_questions": 10,
      "independent_student_answers": 0,
      "later_failures_considered": 6,
      "mixed_source_questions": 12,
      "regressed_question_ids": [
        "generated:local_textcraft_280",
        "generated:local_textcraft_448",
        "generated:local_textcraft_357",
        "generated:cluster_wdt_r1_s4",
        "generated:cluster_wdt_r1_s5",
        "generated:cluster_wdt_r1_s18"
      ],
      "source_questions": 30,
      "upward_sources_skipped_without_teacher_proposals": 1
    },
    "reserved_new_question_positions": 10,
    "main_learning_updates": 1,
    "main_learning_complete": true,
    "model_saved": true,
    "repair": {
      "accepted_by_mode": {
        "prevention": 28,
        "recovery": 13
      },
      "accepted_proposals": 41,
      "optimizer_steps": 1,
      "proposal_count": 57,
      "rejection_reasons": {
        "NO_SUCCESSFUL_RELEVANT_CORRECTION": 1,
        "SUFFIX_DID_NOT_COMPLETE_ORIGINAL_TASK": 15
      },
      "repaired_target_ids": [
        "generated:cluster_wdt_r1_s0",
        "generated:cluster_wdt_r1_s18",
        "generated:cluster_wdt_r1_s4",
        "generated:cluster_wdt_repair_r2_s12",
        "generated:cluster_wdt_repair_r2_s2",
        "generated:cluster_wdt_repair_r2_s3",
        "generated:cluster_wdt_repair_r2_s5",
        "generated:local_textcraft_215",
        "generated:local_textcraft_248",
        "generated:local_textcraft_280",
        "generated:local_textcraft_357",
        "generated:local_textcraft_448",
        "generated:local_textcraft_462",
        "generated:local_textcraft_480",
        "generated:local_textcraft_489",
        "generated:local_textcraft_502"
      ],
      "selected_actions": 65
    },
    "answers_collected_before_second_learning": true,
    "completed_main_learning_rounds_before_collection": 1,
    "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
    "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
    "source_records": {
      "CAPTURE.json": {
        "bytes": 43850,
        "sha256": "d08c7b28c2c8e7f959c8a3b39234c9bcb64dda4a4f2eec6b17ed327d5f989199"
      },
      "complete.json": {
        "bytes": 9348,
        "sha256": "2bad3cd116d11f963246df352be563ae4a5c70501957b5fc37def83e0d193964"
      },
      "collection.json": {
        "bytes": 11662,
        "sha256": "3dae285d9255d10c86bb599c9e832f4d0f0d7d5733506d056f8631789fdb17c5"
      },
      "plan.json": {
        "bytes": 90906,
        "sha256": "ef8012fc618a629be84d4c8ac1a1dbb6990abdc57ad29091d548e1cfad38bcbe"
      },
      "main.json": {
        "bytes": 7690,
        "sha256": "a48a7b6a0280e17a2ccb51f341939339875e2bbab3049bb5482235e302b8a9fb"
      }
    }
  },
  "SCALE50_second_round": {
    "arm": "WDT_SCALE50",
    "round": 2,
    "status": "COMPLETE",
    "questions": 75,
    "answers": 600,
    "seed_questions": 50,
    "seed_answers": 400,
    "seed_successes": 279,
    "review_questions": 25,
    "new_questions": 0,
    "effective_questions": 40,
    "final_policy_id": "20c9a42e5b230319dadf41a42c85c32b7c148dead5b0a32fe9a499c0b098070f",
    "completed_at_utc": "2026-11-02T03:44:15.742210+00:00",
    "completed_at_beijing": "2026-11-02 11:44:15",
    "started_at_utc": "2026-11-02T02:13:21.078609+00:00",
    "complete_round_seconds": 5454.663601,
    "stage_seconds": {
      "plan-round": 4271.307887,
      "train-round": 1183.324984
    },
    "pure_parameter_training_seconds": null,
    "all_correct_questions": 28,
    "all_wrong_questions": 7,
    "effective_new_questions": 0,
    "new_difficulty_by_kind": {},
    "signal_coverage": {
      "failed_questions": 47,
      "first_answer_failed_questions": 27,
      "independent_student_answers": 0,
      "later_failures_considered": 20,
      "mixed_source_questions": 40,
      "regressed_question_ids": [
        "generated:local_textcraft_248",
        "generated:local_textcraft_280",
        "generated:local_textcraft_448",
        "generated:local_textcraft_357",
        "generated:local_textcraft_190",
        "generated:random_r3_textcraft_219",
        "generated:matched_random_r3_textcraft_282",
        "generated:matched_random_r2_textcraft_304",
        "generated:matched_random_r3_textcraft_350",
        "generated:matched_random_r2_textcraft_401",
        "generated:cluster_wdt_scale50_r1_s12",
        "generated:cluster_wdt_scale50_r1_s11",
        "generated:cluster_wdt_scale50_r1_s10",
        "generated:cluster_wdt_scale50_r1_s20",
        "generated:cluster_wdt_scale50_r1_s35",
        "generated:cluster_wdt_scale50_r1_s1"
      ],
      "source_questions": 75,
      "upward_sources_skipped_without_teacher_proposals": 0
    },
    "reserved_new_question_positions": 25,
    "main_learning_updates": 1,
    "main_learning_complete": true,
    "model_saved": true,
    "repair": null,
    "answers_collected_before_second_learning": true,
    "completed_main_learning_rounds_before_collection": 1,
    "answer_meaning": "本轮答卷在第二次学习之前收集，反映第一轮学习后的表现；第二轮模型已经保存。",
    "timing_scope": "完整一轮，包含出题、作答、学习、纠错（如有）、保存和阶段切换。",
    "source_records": {
      "CAPTURE.json": {
        "bytes": 81497,
        "sha256": "91e067ae2740e14e5df137b4df9fbc001f52428b422b65b7df3d1c5be57bd069"
      },
      "complete.json": {
        "bytes": 13748,
        "sha256": "3a3719b6addf1e63cdbfbcbeea0f32e15ee74e41c34011bb198dc13ca9e63ad5"
      },
      "collection.json": {
        "bytes": 21637,
        "sha256": "9d7fc5d35cc7a9980483f1231b04b63ada06f4a01d5b47c916fbac52e6ee613a"
      },
      "plan.json": {
        "bytes": 126053,
        "sha256": "6b3cc77887a2b7503bad20e03b4a6242b8a30475a7e17a641b270d26d2613d34"
      },
      "main.json": {
        "bytes": 13037,
        "sha256": "d94176b7a4f9d90b43afb37b46d8e6394f2ae7192519fd6961ebe70b4e8dcb1f"
      }
    },
    "new_question_issue": {
      "valid_individual_diagnoses": 55,
      "partition_assigned_diagnoses": 54,
      "omitted_diagnosis_id": "D44",
      "generation_candidate_requests": 0,
      "reason": "老师分组漏掉一条诊断，整份分组被拒，55 份有效单题诊断未进入出题。",
      "training_continued_on_existing_questions": true,
      "successful_expansion_this_round": false,
      "evidence": {
        "bytes": 1367166,
        "sha256": "b710a6d99f7f90af62c0fe62af62e5529354b0c5456e415d7e79660b9ef59f39"
      }
    }
  },
  "second_round_extension_summary": "extension-round-2-summary.json",
  "OPT_third_round": {
    "schema_version": "wide_deep_OPT_round3_summary_v1",
    "arm": "WDT_OPT",
    "round": 3,
    "status": "COMPLETE",
    "observed_at_utc": "2026-11-02T04:48:07.580166+00:00",
    "observed_at_beijing": "2026-11-02 12:48:07",
    "completed_at_utc": "2026-11-02T04:46:39.136471+00:00",
    "completed_at_beijing": "2026-11-02 12:46:39",
    "started_at_utc": "2026-11-02T03:44:15.763341+00:00",
    "questions": 44,
    "answers": 352,
    "seed_questions": 20,
    "seed_answers": 160,
    "seed_successes": 120,
    "previous_round_seed_successes": 125,
    "review_questions": 12,
    "new_questions": 12,
    "effective_questions": 17,
    "all_correct_questions": 20,
    "all_wrong_questions": 7,
    "effective_new_questions": 5,
    "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
    "new_difficulty_by_kind": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 2,
        "effective_questions": 0,
        "minimum_actions_delta_range": [
          -11,
          -2
        ],
        "questions": 2,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 1,
        "all_succeeded": 1,
        "effective_questions": 3,
        "minimum_actions_delta_range": [
          1,
          13
        ],
        "questions": 5,
        "required_chain_depth_delta_range": [
          1,
          1
        ]
      },
      "wide": {
        "all_failed": 0,
        "all_succeeded": 3,
        "effective_questions": 2,
        "minimum_actions_delta_range": [
          0,
          1
        ],
        "questions": 5,
        "required_chain_depth_delta_range": [
          0,
          0
        ]
      }
    },
    "signal_coverage": {
      "failed_questions": 18,
      "first_answer_failed_questions": 12,
      "independent_student_answers": 0,
      "later_failures_considered": 6,
      "mixed_source_questions": 12,
      "regressed_question_ids": [
        "generated:local_textcraft_248",
        "generated:local_textcraft_280",
        "generated:local_textcraft_462",
        "generated:local_textcraft_328",
        "generated:cluster_wdt_r1_s18",
        "generated:cluster_wdt_r1_s5",
        "generated:cluster_wdt_opt_r2_s14"
      ],
      "source_questions": 32,
      "upward_sources_skipped_without_teacher_proposals": 1
    },
    "main_learning_updates": 1,
    "main_learning_complete": true,
    "model_saved": true,
    "main_learning_reused": false,
    "new_branch_answers": 352,
    "final_policy_id": "5bac08966225f94f9127a524698cf56b33eb41f5ee47ffdbe4bdfd399ed4db45",
    "answers_collected_before_third_learning": true,
    "completed_main_learning_rounds_before_collection": 2,
    "answer_meaning": "第三轮答卷在第三次学习前收集，反映前两轮学习后的表现；本轮模型已保存。",
    "effect_interpretation": "共同20道种子由上一轮125/160变为120/160，当前尚无稳定原题收益；第三次学习的效果待共同30题测试。",
    "final_common30_test_completed": false,
    "complete_round_seconds": 3743.37313,
    "stage_seconds": {
      "plan-round": 2582.338237,
      "train-round": 1161.00745
    },
    "pure_parameter_training_seconds": null,
    "timing_scope": "完整一轮，含出题、作答、学习、保存和阶段切换。",
    "question_results": [
      {
        "task_id": "generated:local_textcraft_188",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_215",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_248",
        "section": "seed",
        "kind": "root",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_273",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_280",
        "section": "seed",
        "kind": "root",
        "successes": 3,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_42",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_448",
        "section": "seed",
        "kind": "root",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_462",
        "section": "seed",
        "kind": "root",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_477",
        "section": "seed",
        "kind": "root",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_480",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_489",
        "section": "seed",
        "kind": "root",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_502",
        "section": "seed",
        "kind": "root",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_517",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_328",
        "section": "seed",
        "kind": "root",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_357",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_100",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_72",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_45",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_190",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_355",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_r1_s18",
        "section": "review",
        "kind": "wide",
        "successes": 6,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s4",
        "section": "review",
        "kind": "wide",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_r1_s1",
        "section": "review",
        "kind": "deep_down",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s5",
        "section": "review",
        "kind": "deep_down",
        "successes": 1,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s18",
        "section": "review",
        "kind": "deep_up",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s0",
        "section": "review",
        "kind": "wide",
        "successes": 6,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s17",
        "section": "review",
        "kind": "deep_down",
        "successes": 6,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s11",
        "section": "review",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s14",
        "section": "review",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s26",
        "section": "review",
        "kind": "deep_down",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s12",
        "section": "review",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r2_s2",
        "section": "review",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s0",
        "section": "new",
        "kind": "wide",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s3",
        "section": "new",
        "kind": "deep_up",
        "successes": 4,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s4",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s15",
        "section": "new",
        "kind": "deep_up",
        "successes": 1,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s18",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s5",
        "section": "new",
        "kind": "deep_down",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s6",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s7",
        "section": "new",
        "kind": "deep_up",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s20",
        "section": "new",
        "kind": "wide",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s19",
        "section": "new",
        "kind": "deep_up",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s9",
        "section": "new",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_opt_r3_s11",
        "section": "new",
        "kind": "deep_down",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      }
    ],
    "source_records": {
      "CAPTURE.json": {
        "bytes": 56218,
        "sha256": "c8d714da49fb7486e237bdaa098f1ebf44cd83326e782d8e371cb3fd22a8f2ee"
      },
      "complete.json": {
        "bytes": 9013,
        "sha256": "4ac1899cd2b0324106031cb4a65764ed1f173599efa2794ff06ee3b798cd8510"
      },
      "collection.json": {
        "bytes": 12803,
        "sha256": "922c6d8d09c7ee673ebb019058dcf4ca9b7a78413292328fb993fba8236e7a55"
      },
      "plan.json": {
        "bytes": 97862,
        "sha256": "329e33b2eca23c3cbd64ec4c8059a356eb96fe7b82348c7043311e6e61ba9d0a"
      },
      "main.json": {
        "bytes": 8309,
        "sha256": "0c127a4cef51b8b980f146cd750fdaa55a93c09672b1396dfb0cc95591bb31e5"
      },
      "main_model/COMPLETE.json": {
        "bytes": 337,
        "sha256": "a554c4e4fa5036b7cd2bdf652584005eab875a59947c9c7276e1c94da806b771"
      },
      "timing/WDT_OPT_plan-round_3_attempt_01.started.json": {
        "bytes": 1126,
        "sha256": "bc14a071791f28522102c815cce1a0c7142d1a6f5996736461dc6145ca9f5e3c"
      },
      "timing/WDT_OPT_train-round_3_attempt_01.exit.json": {
        "bytes": 178,
        "sha256": "197014fad7e059c45549f2f94905a4c2332dcc29e400a688d6e8cacda4bd9e70"
      }
    }
  },
  "OPT_training_complete": true,
  "third_round_OPT_summary": "opt-round-3-summary.json",
  "REPAIR_third_round": {
    "schema_version": "wide_deep_REPAIR_round3_summary_v1",
    "arm": "WDT_REPAIR",
    "round": 3,
    "status": "COMPLETE",
    "observed_at_utc": "2026-11-02T06:28:20.285682+00:00",
    "observed_at_beijing": "2026-11-02 14:28:20",
    "started_at_utc": "2026-11-02T04:46:39.165028+00:00",
    "completed_at_utc": "2026-11-02T06:22:35.663424+00:00",
    "completed_at_beijing": "2026-11-02 14:22:35",
    "questions": 44,
    "answers": 352,
    "seed_questions": 20,
    "seed_answers": 160,
    "seed_successes": 118,
    "previous_round_seed_successes": 125,
    "common_old_questions": 32,
    "common_old_answers": 256,
    "common_old_successes_previous": 159,
    "common_old_successes_current": 164,
    "review_questions": 12,
    "new_questions": 12,
    "effective_questions": 18,
    "all_correct_questions": 17,
    "all_wrong_questions": 9,
    "effective_new_questions": 4,
    "effective_questions_definition": "同一道题的八份回答有对有错，能提供本轮主练的成败比较。",
    "new_difficulty_by_kind": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 1,
        "effective_questions": 1,
        "minimum_actions_delta_range": [
          -5,
          -4
        ],
        "questions": 2,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 1,
        "all_succeeded": 0,
        "effective_questions": 3,
        "minimum_actions_delta_range": [
          2,
          4
        ],
        "questions": 4,
        "required_chain_depth_delta_range": [
          1,
          1
        ]
      },
      "wide": {
        "all_failed": 2,
        "all_succeeded": 4,
        "effective_questions": 0,
        "minimum_actions_delta_range": [
          -1,
          1
        ],
        "questions": 6,
        "required_chain_depth_delta_range": [
          -1,
          0
        ]
      }
    },
    "signal_coverage": {
      "failed_questions": 20,
      "first_answer_failed_questions": 10,
      "independent_student_answers": 0,
      "later_failures_considered": 10,
      "mixed_source_questions": 14,
      "regressed_question_ids": [
        "generated:local_textcraft_188",
        "generated:local_textcraft_248",
        "generated:local_textcraft_280",
        "generated:local_textcraft_462",
        "generated:local_textcraft_502",
        "generated:local_textcraft_517"
      ],
      "source_questions": 32,
      "upward_sources_skipped_without_teacher_proposals": 1
    },
    "main_learning_updates": 1,
    "repair_learning_updates": 1,
    "new_learning_updates_completed": 2,
    "main_learning_complete": true,
    "repair_learning_complete": true,
    "main_model_saved": true,
    "repair_model_saved": true,
    "model_saved": true,
    "main_learning_reused": false,
    "new_branch_answers": 352,
    "collection_policy_id": "64bba71af88c1427cc676e23f26a83f3f6f70ead23698dae58c93bd283d64fe7",
    "main_policy_id": "f445a3e252ec14df358c1fc37bb26fd7c0756cce8034ad607926b0642b7330a7",
    "final_policy_id": "aeca99455724c3b1855ef6b0ed30ea2fce7dd20dfd01f7a907d386b0c2667ce3",
    "previous_round_final_model_matches_collection_model": true,
    "answers_collected_before_third_learning": true,
    "completed_main_learning_rounds_before_collection": 2,
    "answer_meaning": "第三轮答卷在本轮主练和纠错前收集，反映第二轮最终模型。第三轮最终效果待共同30题测试。",
    "effect_interpretation": "固定20道种子由125/160降至118/160；共同32道旧题由159/256升至164/256，两个范围的变化方向不同。",
    "repair": {
      "accepted_by_mode": {
        "prevention": 31,
        "recovery": 15
      },
      "accepted_proposals": 46,
      "optimizer_steps": 1,
      "proposal_count": 69,
      "rejection_reasons": {
        "NO_SUCCESSFUL_RELEVANT_CORRECTION": 2,
        "SUFFIX_DID_NOT_COMPLETE_ORIGINAL_TASK": 21
      },
      "repaired_target_ids": [
        "generated:cluster_wdt_r1_s0",
        "generated:cluster_wdt_r1_s18",
        "generated:cluster_wdt_repair_r2_s12",
        "generated:cluster_wdt_repair_r2_s2",
        "generated:cluster_wdt_repair_r2_s3",
        "generated:cluster_wdt_repair_r2_s5",
        "generated:cluster_wdt_repair_r3_s21",
        "generated:cluster_wdt_repair_r3_s23",
        "generated:cluster_wdt_repair_r3_s5",
        "generated:cluster_wdt_repair_r3_s7",
        "generated:local_textcraft_188",
        "generated:local_textcraft_248",
        "generated:local_textcraft_280",
        "generated:local_textcraft_448",
        "generated:local_textcraft_462",
        "generated:local_textcraft_489",
        "generated:local_textcraft_502",
        "generated:local_textcraft_517"
      ],
      "selected_actions": 73,
      "original_target_count": 23,
      "original_goal_completed_proposals": 47,
      "accepted_target_count": 18,
      "completed_original_goal_without_accepted_actions": 1,
      "selected_correction_tokens": 1055,
      "original_target_cases": [
        {
          "task_id": "generated:local_textcraft_477",
          "label": "书架",
          "proposals": 3,
          "original_goal_completed_proposals": 0,
          "accepted_proposals": 0
        },
        {
          "task_id": "generated:local_textcraft_489",
          "label": "完整红旗",
          "proposals": 3,
          "original_goal_completed_proposals": 3,
          "accepted_proposals": 3
        },
        {
          "task_id": "generated:local_textcraft_188",
          "label": "安山岩",
          "proposals": 3,
          "original_goal_completed_proposals": 3,
          "accepted_proposals": 2
        }
      ],
      "original_goal_completion_definition": "原记录 judge_status 为 PASS；与是否存在合格学习动作分别统计。",
      "accepted_proposals_definition": "完成原目标并通过改正动作筛选的方案。",
      "selected_actions_definition": "去掉重复项后，实际用于纠错学习的合格改正动作数。",
      "student_independent_success_after_learning_verified": false
    },
    "final_common30_test_completed": false,
    "complete_round_seconds": 5756.498396,
    "stage_seconds": {
      "plan-round": 2566.931565,
      "train-round": 3189.53828
    },
    "pure_parameter_training_seconds": null,
    "timing_scope": "完整一轮，含出题、作答、主练、纠错、保存和阶段切换。",
    "stage_observation_intervals": {
      "remaining_answers_and_preparation_seconds_interval": [
        696.882401,
        743.968748
      ],
      "main_learning_plus_saving_seconds_interval": [
        450.458368,
        543.847943
      ],
      "repair_material_generation_and_validation_seconds_interval": [
        1892.11893,
        1977.927737
      ],
      "repair_learning_saving_and_exit_seconds_interval": [
        17.183427,
        56.689006
      ]
    },
    "stage_observation_interval_scope": "相邻观察确定的时间范围，不是纯模型计算时间；区间上限不能直接相加。",
    "question_results": [
      {
        "task_id": "generated:local_textcraft_188",
        "section": "seed",
        "kind": "root",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_215",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_248",
        "section": "seed",
        "kind": "root",
        "successes": 3,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_273",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_280",
        "section": "seed",
        "kind": "root",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_42",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_448",
        "section": "seed",
        "kind": "root",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_462",
        "section": "seed",
        "kind": "root",
        "successes": 3,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_477",
        "section": "seed",
        "kind": "root",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_480",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_489",
        "section": "seed",
        "kind": "root",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_502",
        "section": "seed",
        "kind": "root",
        "successes": 4,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_517",
        "section": "seed",
        "kind": "root",
        "successes": 6,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:local_textcraft_328",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_357",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_100",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_72",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_45",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_190",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:local_textcraft_355",
        "section": "seed",
        "kind": "root",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_r1_s18",
        "section": "review",
        "kind": "wide",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s4",
        "section": "review",
        "kind": "wide",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_r1_s5",
        "section": "review",
        "kind": "deep_down",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s10",
        "section": "review",
        "kind": "deep_down",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s5",
        "section": "review",
        "kind": "wide",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_r1_s1",
        "section": "review",
        "kind": "deep_down",
        "successes": 4,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s17",
        "section": "review",
        "kind": "deep_down",
        "successes": 6,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s3",
        "section": "review",
        "kind": "wide",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_r1_s0",
        "section": "review",
        "kind": "wide",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s12",
        "section": "review",
        "kind": "deep_up",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s14",
        "section": "review",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r2_s2",
        "section": "review",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s0",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s1",
        "section": "new",
        "kind": "deep_down",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s16",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s17",
        "section": "new",
        "kind": "deep_down",
        "successes": 7,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s4",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s5",
        "section": "new",
        "kind": "deep_up",
        "successes": 5,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s20",
        "section": "new",
        "kind": "wide",
        "successes": 8,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s21",
        "section": "new",
        "kind": "deep_up",
        "successes": 2,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s6",
        "section": "new",
        "kind": "wide",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s7",
        "section": "new",
        "kind": "deep_up",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s22",
        "section": "new",
        "kind": "wide",
        "successes": 0,
        "attempts": 8,
        "has_signal": false,
        "included_in_update": false
      },
      {
        "task_id": "generated:cluster_wdt_repair_r3_s23",
        "section": "new",
        "kind": "deep_up",
        "successes": 3,
        "attempts": 8,
        "has_signal": true,
        "included_in_update": true
      }
    ],
    "source_records": {
      "CAPTURE.json": {
        "bytes": 58292,
        "sha256": "481c8837533874f3fcc9d252a3779276d4967bba13aaf908273090373464d3c0"
      },
      "complete.json": {
        "bytes": 9988,
        "sha256": "94889594a144983fd0a0bd53ce66a79f38c72aeb7b8218d4be6868492535af1e"
      },
      "collection.json": {
        "bytes": 12803,
        "sha256": "05ccda0149e1ca38cabe05bb2db12793d5cda60a5853ef2ecaa99a684e720b6f"
      },
      "plan.json": {
        "bytes": 101855,
        "sha256": "cf4a8aac703816624a1ad4bc09cc51656228a6ff920b5947ab9f1286b220e0c8"
      },
      "main.json": {
        "bytes": 8330,
        "sha256": "80ea9fbc1874462bb70cd08e5b134e9ead463148db9d672c67b66e6020b6a759"
      },
      "repair.json": {
        "bytes": 952,
        "sha256": "a9aac9f467d3dd314e804ea9a0205150d5607d8ca7928fd4b9ec097e4a6008c5"
      },
      "repair_summary.json": {
        "bytes": 116040,
        "sha256": "3c50f288bb48e2ec1440abfa4794cfc601f79c6ba574a786780a675d4dbab041"
      },
      "main_model/COMPLETE.json": {
        "bytes": 337,
        "sha256": "9af78ea8fced779502f596ff16b3076f12dc05de0324b1110ac2a350e4649d99"
      },
      "repair_model/COMPLETE.json": {
        "bytes": 337,
        "sha256": "938db3362a51c2379bf21664056a7281fe2cfe67a92f2a47b8dad82bd397ea49"
      },
      "timing/WDT_REPAIR_plan-round_3_attempt_01.started.json": {
        "bytes": 1141,
        "sha256": "b9c492335e3fbf15cb8a3a29143d5b1d0416666fd571fce252f7f02e72208b5c"
      },
      "timing/WDT_REPAIR_plan-round_3_attempt_01.exit.json": {
        "bytes": 180,
        "sha256": "5dc4349ad5667c350da7433bf1d232e41b7b24498a636f6de6f52b4c60d5d19f"
      },
      "timing/WDT_REPAIR_train-round_3_attempt_01.started.json": {
        "bytes": 1148,
        "sha256": "76bf2a00a9f18084a8d6b25eb69acf9225e24c4cf8e08aaaffe85c9dc412edc5"
      },
      "timing/WDT_REPAIR_train-round_3_attempt_01.exit.json": {
        "bytes": 181,
        "sha256": "e6fe2939a08c4705ad4a143a9e4599d52f6d6b8a92c86fae887284913c6153e1"
      }
    },
    "parent_analysis": {
      "bytes": 8575,
      "sha256": "559e0778b28fff5ece438ae2b57f577dfd0d50c06e2ffe4829c427fd68a9e80b"
    }
  },
  "REPAIR_training_complete": true,
  "third_round_REPAIR_summary": "repair-round-3-summary.json",
  "observed_time_precision": "minute",
  "all_main_training_complete": true,
  "final_evaluation_started": true,
  "SCALE50_third_round": {
    "arm": "WDT_SCALE50",
    "round": 3,
    "status": "COMPLETE",
    "seed_questions": 50,
    "review_questions": 25,
    "new_questions": 9,
    "questions": 84,
    "answers": 672,
    "answers_per_question": 8,
    "effective_questions": 40,
    "effective_new_questions": 4,
    "all_succeeded_questions": 33,
    "all_failed_questions": 11,
    "new_main_learning_updates": 1,
    "final_model_saved": true,
    "repair_enabled": false,
    "new_question_breakdown": {
      "deep_down": {
        "all_failed": 0,
        "all_succeeded": 1,
        "effective_questions": 0,
        "minimum_actions_delta_range": [
          -10,
          -10
        ],
        "questions": 1,
        "required_chain_depth_delta_range": [
          -1,
          -1
        ]
      },
      "deep_up": {
        "all_failed": 0,
        "all_succeeded": 0,
        "effective_questions": 2,
        "minimum_actions_delta_range": [
          1,
          4
        ],
        "questions": 2,
        "required_chain_depth_delta_range": [
          1,
          1
        ]
      },
      "wide": {
        "all_failed": 1,
        "all_succeeded": 3,
        "effective_questions": 2,
        "minimum_actions_delta_range": [
          -3,
          1
        ],
        "questions": 6,
        "required_chain_depth_delta_range": [
          -1,
          1
        ]
      }
    },
    "seed_observations": {
      "questions": 50,
      "previous_round": 2,
      "current_round": 3,
      "previous_successes": 279,
      "previous_answers": 400,
      "current_successes": 279,
      "current_answers": 400,
      "improved_questions": 12,
      "regressed_questions": 17,
      "unchanged_questions": 21,
      "recovered_from_all_failed": 0,
      "regressed_to_all_failed": 2
    },
    "common_old_task_observations": {
      "questions": 75,
      "previous_round": 2,
      "current_round": 3,
      "previous_successes": 402,
      "previous_answers": 600,
      "current_successes": 407,
      "current_answers": 600,
      "improved_questions": 20,
      "regressed_questions": 22,
      "unchanged_questions": 33,
      "recovered_from_all_failed": 1,
      "regressed_to_all_failed": 4
    },
    "observation_interpretation": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
    "start_utc": "2026-11-02T06:22:35.695521+00:00",
    "end_utc": "2026-11-02T08:10:21.268705+00:00",
    "start_beijing": "2026-11-02 14:22:35",
    "end_beijing": "2026-11-02 16:10:21",
    "full_elapsed_seconds": 6465.573184,
    "stage_sum_seconds": 6465.546611000001,
    "interstage_seconds": 0.026573,
    "recorded_stage_seconds": {
      "plan-round": 4874.832532,
      "train-round": 1590.714079
    },
    "pure_parameter_training_seconds": null,
    "elapsed_scope": "整轮用时包含出题、作答、学习、保存及环节交接；后半段也不是纯模型调整时间。",
    "schema_version": "wide_deep_SCALE50_round3_summary_v1",
    "completed_at_utc": "2026-11-02T08:10:21.268705+00:00",
    "completed_at_beijing": "2026-11-02 16:10:21",
    "started_at_utc": "2026-11-02T06:22:35.695521+00:00",
    "complete_round_seconds": 6465.573184,
    "model_saved": true,
    "main_learning_complete": true,
    "main_learning_updates": 1,
    "new_learning_updates_completed": 1,
    "answers_collected_before_third_learning": true,
    "completed_main_learning_rounds_before_collection": 2,
    "answer_meaning": "第二轮与第三轮作答的比较；第三轮答卷来自第二轮最终模型。第三轮训练后的最终表现由共同测试衡量。",
    "final_common30_test_completed": false,
    "source_public_summary": "scale50-round-3-summary.json"
  },
  "SCALE50_training_complete": true,
  "third_round_SCALE50_summary": "scale50-round-3-summary.json",
  "main_training_completed_at_beijing": "2026-11-02 16:10:21",
  "later_test_observation": {
    "schema": "main_session_test_progress_public_update_v1",
    "source": "experiment_main_session_handoff",
    "as_of_beijing": "2026-11-02 17:29",
    "observed_time_precision": "minute",
    "training": {
      "completed_rounds": 21,
      "required_rounds": 21,
      "saved_models_unchanged": true
    },
    "N": {
      "status": "COMPLETE",
      "answers_saved": 90,
      "answers_required": 90,
      "complete": true
    },
    "WT": {
      "status": "INCOMPLETE_NEEDS_RESUME",
      "answers_saved": 84,
      "answers_required": 90,
      "answers_missing": 6,
      "complete": false,
      "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
      "input_length_limit": 4096,
      "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
      "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
      "recovery_owner": "experiment_main_session",
      "previous_public_record": "wt-test-interruption-1714.json"
    },
    "DT": {
      "status": "INCOMPLETE_NEEDS_RESUME",
      "answers_saved": 84,
      "answers_required": 90,
      "answers_missing": 6,
      "complete": false,
      "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
      "input_length_limit": 4096,
      "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
      "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
      "recovery_owner": "experiment_main_session"
    },
    "WDT": {
      "status": "RUNNING",
      "answers_saved": 44,
      "answers_required": 90,
      "complete": false
    },
    "completed_test_models": [
      "N"
    ],
    "main_answers_saved_from_reported_counts": 302,
    "main_answers_required": 720,
    "main_answer_total_basis": "已完成基础模型90份 + WT已存84份 + DT已存84份 + WDT已存44份。",
    "auxiliary_last_known_answers_saved": 0,
    "auxiliary_answers_required": 180,
    "auxiliary_last_known_observation_beijing": "2026-11-02 16:37:20",
    "completion_estimate": {
      "status": "RECALCULATION_PENDING_AFTER_RECOVERY",
      "current_estimate": null,
      "previous_estimate_valid_for_current_schedule": false,
      "display": "补测恢复后重新估计"
    }
  },
  "main_test_models_completed": 1,
  "main_test_completed_results": [
    {
      "label": "N",
      "complete": true,
      "summary_matches_scope": true,
      "questions": 30,
      "answers": 90,
      "sampled_successes": 27,
      "sampled_attempts": 60,
      "sampled_rate": 0.45,
      "greedy_successes": 15,
      "greedy_attempts": 30,
      "greedy_rate": 0.5
    }
  ],
  "main_test_partial_results_available": true,
  "main_test_result_summary": "base-test-summary.json",
  "auxiliary_test_observed_at_beijing": "2026-11-02 16:37:20",
  "main_test_answers_count_basis": "已完成基础模型90份 + WT已存84份 + DT已存84份 + WDT已存44份。",
  "test_interruptions": [
    {
      "label": "WT",
      "status": "INCOMPLETE_NEEDS_RESUME",
      "answers_saved": 84,
      "answers_required": 90,
      "answers_missing": 6,
      "complete": false,
      "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
      "input_length_limit": 4096,
      "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
      "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
      "recovery_owner": "experiment_main_session",
      "previous_public_record": "wt-test-interruption-1714.json"
    },
    {
      "label": "DT",
      "status": "INCOMPLETE_NEEDS_RESUME",
      "answers_saved": 84,
      "answers_required": 90,
      "answers_missing": 6,
      "complete": false,
      "public_reason": "较长题目的完整作答历史超过输入长度限制，测试中断。",
      "input_length_limit": 4096,
      "recovery_status": "LOCAL_FIX_COMPLETE_RESUME_PACKAGE_PREPARING",
      "planned_recovery_scope": "保留已完成答卷，只补缺失答卷，不重新训练。",
      "recovery_owner": "experiment_main_session"
    }
  ],
  "completion_estimate_available": false
}
