P论文工作台最新版本 ↗
完整资料 / R5

500共同环境最终汇总(纠正后)

paper-rl/results/common_profile_runtime_coverage_final_20260930.json

result观察/版本:2026-10-01 纠正31,698 bytes下载完整原文直接读取
{
  "observed_at": "2026-09-30T15:04:58.962898+08:00",
  "scope": "All500 common-profile tasks have been exercised across retained waves; unresolved and mismatches remain explicit. No model-effect claim.",
  "common_profile_tasks": 500,
  "counts": {
    "controls": 1000,
    "raw_pass": 995,
    "strict_pass": 862,
    "report_binary": 887,
    "unresolved": 113,
    "partial_zeros": 20,
    "control_expectation_mismatches": 5,
    "binary_decisions_matching_control_expectation": 882
  },
  "cohorts": [
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2",
      "tasks": 8,
      "controls": 16,
      "raw_pass": 16,
      "strict_pass": 16,
      "report_binary": 16,
      "unresolved": 0,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 48,
      "report_binary": 48,
      "unresolved": 0,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 40,
      "report_binary": 41,
      "unresolved": 7,
      "partial_zeros": 1
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 43,
      "report_binary": 43,
      "unresolved": 5,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5",
      "tasks": 7,
      "controls": 14,
      "raw_pass": 14,
      "strict_pass": 10,
      "report_binary": 10,
      "unresolved": 4,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation01-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 47,
      "strict_pass": 44,
      "report_binary": 45,
      "unresolved": 3,
      "control_expectation_mismatches": 1
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation02-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 40,
      "report_binary": 42,
      "unresolved": 6,
      "partial_zeros": 2
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 39,
      "report_binary": 39,
      "unresolved": 9,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 33,
      "report_binary": 36,
      "unresolved": 12,
      "partial_zeros": 3
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 35,
      "report_binary": 38,
      "unresolved": 10,
      "partial_zeros": 3
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 36,
      "report_binary": 39,
      "unresolved": 9,
      "partial_zeros": 3
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 38,
      "report_binary": 40,
      "unresolved": 8,
      "partial_zeros": 2
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 37,
      "report_binary": 39,
      "unresolved": 9,
      "partial_zeros": 2
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 38,
      "report_binary": 38,
      "unresolved": 10,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 38,
      "report_binary": 40,
      "unresolved": 8,
      "partial_zeros": 2
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 47,
      "strict_pass": 42,
      "report_binary": 44,
      "unresolved": 4,
      "partial_zeros": 1,
      "control_expectation_mismatches": 1
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 47,
      "strict_pass": 47,
      "report_binary": 48,
      "unresolved": 0,
      "partial_zeros": 0,
      "control_expectation_mismatches": 1
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 46,
      "strict_pass": 46,
      "report_binary": 48,
      "unresolved": 0,
      "partial_zeros": 0,
      "control_expectation_mismatches": 2
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 45,
      "report_binary": 45,
      "unresolved": 3,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 48,
      "report_binary": 48,
      "unresolved": 0,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 46,
      "report_binary": 47,
      "unresolved": 1,
      "partial_zeros": 1
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930",
      "tasks": 24,
      "controls": 48,
      "raw_pass": 48,
      "strict_pass": 45,
      "report_binary": 45,
      "unresolved": 3,
      "partial_zeros": 0
    },
    {
      "run": "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930",
      "tasks": 5,
      "controls": 10,
      "raw_pass": 10,
      "strict_pass": 8,
      "report_binary": 8,
      "unresolved": 2,
      "partial_zeros": 0
    }
  ],
  "independent_native_replay_scope": "Prior16-control integration plus selected earlier anomaly slices and Django10097 old/new4controls; not all270controls cold-reviewed.",
  "remaining_common_profile_tasks_without_new_runtime_controls": 0,
  "execution_ready": false,
  "adopted": false,
  "input_sha256": {
    "rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2/plan.json": "e43c210c0d15b3d0ff08e63a25182bea45fe4bdcce4d6a2443ee37f9e8b88c60",
    "rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2/records.jsonl": "bce8a90485533394d40e9dde3767f15b4be05ecddd4cc7ded29ba5acc1b577bd",
    "rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2/raw-audit.json": "ba022b62c2ab75d99ffe9ac60e3f912081a164313a1b7e57b5a7f3d599d61995",
    "rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2/observations-audit.json": "1af6dd7e5a6dd686b8a3d0a5d40fe1bfc87c08f6f03d93b9731a1c1d8bf504c0",
    "rl/runs/rl-paper-20260927/fixtures/evalctl-common-offline-integration-v2/report-diagnostic-v1.json": "d15371577e3e77afd9513892639fdd7489823fd05774c504ad64242007e313f0",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5/plan.json": "a9fd51100e2ef9a5fbbc193bfe6b5dc665b8137f47d6d1a10f52d583629afc93",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5/records.jsonl": "57a5e2571e764874d92eaa18368dfdd2778aa6dbbb26f3e333be1ea1bf4f126a",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5/raw-audit.json": "84de28ac3e85fb6473feffded86121f800d6e07da3b9995a76ed738512f4fc4a",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5/observations-audit.json": "2b9d67af898e4e45e4f894b8e2e593547974a056316afd4a4d44776281a59c9e",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch15-common-v5/report-diagnostic-v1.json": "2de40db5d2409c05075c11a6a8ac8c659fee05c0840c5199b3d3e6b48fd8d015",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5/plan.json": "bc95d2781e3718e5e9cc708a0c04551b3f1fb35b57f590f276fc5837c97984b2",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5/records.jsonl": "bf48de0d1d35f20b2ff7f923572f2a0ef78d14592c4c470b285c351bb2d2d732",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5/raw-audit.json": "b42731d33c35cf737e57f327f467e198dd824a6d6467d8911db5dfb22d31b487",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5/observations-audit.json": "b538af9d7e9b7d91b01e56b52c075b0c7593ba19e729afab1e769527986711c6",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch16-common-v5/report-diagnostic-v1.json": "d5aa4d9a6b47cf71d1526c9d6c0af441010f3a59cb647f826534c525cff3b236",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5/plan.json": "8771987b5cdb6aaa3b4379d4d534940943809862963588582fa30e24fdf1c0ec",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5/records.jsonl": "49f216e4d8adacf04bda9d2dea9e7736240a315805be937fc3ebc76b6f0926fc",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5/raw-audit.json": "612d9f5e2bd205e9c53614bd4226b3131ebd31e8a1c9a3cb51cb96757ec3bf28",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5/observations-audit.json": "5404848fef03a0812c9fe3af0353ab28a2c66ffb9c72403516162c90ad0a04e5",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch17-common-v5/report-diagnostic-v1.json": "9d0cd51353510bde51f76ed9d1a135a069532b7604ac835ff722b6fd0ea6fba0",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5/plan.json": "7bf7126151baa95bf0ae906b64ebf347fea922b901984b8c3572b02739818b93",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5/records.jsonl": "37cbf52ff3f200bb24b0804c04af96fe6c8137452d55147c06b1e19f03c65944",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5/raw-audit.json": "7e0da68ccd51e054309cb371d80831a90defe7ecbc61f10c0b731b48a5660d7d",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5/observations-audit.json": "6298fc17601e3fa8b49aeae9fb6d9790851561f3b054562ef9e150f071898dc4",
    "rl/runs/rl-paper-20260927/fixtures/eval500-newprofiles-batch18-common-v5/report-diagnostic-v1.json": "ac357c8661f8d213125ca4954fe8c1d49b385b30e61e2aa4e29df7ceb212ad38",
    "paper-rl/results/common_profile_runtime_coverage_20260928.json": "ddd991e701f00b815086d4343ef0ae12f8ce9c96d4d5b90799704e68de11f09a",
    "paper-rl/results/common_validation01_completed_20260928.json": "cdfa2c40a81de9664d495e7386617f74756f4d75812f5997f0cfe7de8cd7c4dc",
    "rl/runs/rl-paper-20260927/fixtures/eval500-common-validation01-20260928/plan.json": "a7b5d1f34df1bd7e14ce6af0c7cb782a70af023bb0274a3da308843c95afa5a0",
    "rl/runs/rl-paper-20260927/fixtures/eval500-common-validation01-20260928/records.jsonl": "5f2fac8149927c680cb9dbc8eaf988354742d531204d17dd65a39dbe41a526eb",
    "paper-rl/results/common_profile_runtime_coverage_v2_20260928.json": "508dfa6fa2ec66ae9e6829f39c9fd9667e749b7da093a683bdc8ebe765d29246",
    "paper-rl/results/common_validation02_completed_20260928.json": "3e02c4e2da071f64864cda119c49071e05b0c93c2fe9a5a26bf1dc99f4bb83a0",
    "rl/runs/rl-paper-20260927/fixtures/eval500-common-validation02-20260928/plan.json": "ed57771dcee2e7fde2407c8f4cbb995ce4998c4a4d1059a07a6cd1bcfe0521d5",
    "rl/runs/rl-paper-20260927/fixtures/eval500-common-validation02-20260928/records.jsonl": "307cd39adf831a49357f68df0235f824140e2bc653c821a940309572453ea10b",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/plan.json": "507623f08b73fa5a92d09cdbadd9c6776bcfa4df26db9825a616d49ad1853851",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/records.jsonl": "7a9ace6e5d8bd00a8866f07917870165cfd5115c8d0b81b370afacc4fa952d71",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/raw-audit-v1.json": "2223b63aad7948ea48d6b8e4aea8493235504979f5f7fcb87765ffa733f0f925",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/observations-audit-v1.json": "1bf4079c96e681317a6faf86008a410f3080a8275eb5713dd87159484b125ea7",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/report-diagnostic-v1.json": "355a44382566a5fa9a8ee22cc0135075e903f9779e5daf85e0c3d2ca841cc2cd",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation03-20260928/exit-projects-clear-v1.json": "90f5c35e3888d1694bc0ce57751767180cf8b0c43d81eb2d5b69c719d16613a1",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/plan.json": "58fcf7ab3eb17a8c312e71e338b17d60a9f888e52eca7862089c780a6590be65",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/records.jsonl": "cc587166947e69ce2944a268364f7abd5963a3e0055d17b2a2b150688550e05a",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/raw-audit-v1.json": "193f3c520f9fc1d683d946085f3eeb6f594fdc825c362f6ca2aac84fb5a6511d",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/observations-audit-v1.json": "c76f9d6e8f6dadc38d54b017d5a4f702fbad3e7e381cf943b3643cc27b0e4ce6",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/report-diagnostic-v1.json": "25e8c8af3470fa5a1d1be0a63b0039064ab0676942808151ef3d7ef1665ef0c3",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation04-20260928/exit-projects-clear-v1.json": "e8dfff79a96dbf2f73d0dc6568e9ca1cb5310ecb6abb333b0cda4e366e60661d",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/plan.json": "50cf03a4cb32a32e6c4b4048e9763ed9435a33c864aeace643d332010efdb1cc",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/records.jsonl": "04a98784743c156c7a0b0f33a0cd8d93a11849505efb239c2c2597c8f25813cc",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/raw-audit-v1.json": "5df2e6dea6066ad9ac23f82a3b6315a32993bab273b5469a2b6ff8108580abd2",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/observations-audit-v1.json": "f2eec165c9e995bf740513716ffa145472cd2b7b6bb9030471a4efe1d24e3a64",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/report-diagnostic-v1.json": "5570ccd91f1a4ec1c3fb5764f31c5908629b867f9ddcc18e4b978da05715ca4c",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation05-20260928/exit-projects-clear-v1.json": "564567c10f6466057eeaee097ae8dc86bda311892932a139d02b8d2cf5161d1c",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/plan.json": "ec508beabeb6cbc446c2963ac6160a023c2735d7fcae22bb49470991cdc9e001",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/records.jsonl": "901fce48e2e2ee3c7eee1b502ac97295d5dd4b1ea826ac7d22751920eb0f90bd",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/raw-audit-v1.json": "4df093764dc236cb4b3e23c0a538e39ff3931caa8e9b4c6184b80204a737df2a",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/observations-audit-v1.json": "eef24223ee9b6e10807bbf5c03c9961721b4cb18c287d9707f63418eecd420b7",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/report-diagnostic-v1.json": "eadeab154420623fbd5fc8473b61f3e500373d65913bae0999aaf737a451f21e",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation06-20260928/exit-projects-clear-v1.json": "76dec06228947559a8de9ecff386f86632ddf92518afe09fbfc0905803c2b5bd",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/plan.json": "be6b165a87ae1d322106e960e25c6acb6af86595a3e6f073bcc8f1b2477a2c48",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/records.jsonl": "22a7bb091790e17d74411166090d7fa5309dc00d7927a0ac8a8da7aa94575460",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/raw-audit-v1.json": "10a69dec067238103fba3a929260b74e32fe7e87033ad5e15a9995a64a7b8772",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/observations-audit-v1.json": "573343734ff966d351bf1854fe8944eb6a53497d14b64db34fb65b62fcd47836",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/report-diagnostic-v1.json": "ca511c3c32999f76fe1f22995c5ca70217873183aab590c40a0e13dd96192b70",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation07-20260928/exit-projects-clear-v1.json": "192c6c247e7b754ff83f8986d0b533fbe2f367227b4cc87fa19c45c3c9568e6d",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/plan.json": "a6b1ded5a046b0cf9f39023c686193b6bbc05857bf848fbd88b9ba99b469a9f3",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/records.jsonl": "ed7338dcb1ac201abf57ab719f20c2061f88295b85a115a6d82dc247241b6898",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/raw-audit-v1.json": "523cced028b8ec5d6d552ee75ec2cb56ba9918637a093e1b71586c653f205e20",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/observations-audit-v1.json": "f7b755c9212839a80b8028b74842d18577e6d88cf2d6f07b9749c9c506421fca",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/report-diagnostic-v1.json": "9da2dbd8bf4a85104b44ab2d461a5746d779b7e6bda5459362c26928d73e87bc",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation08-20260928/exit-projects-clear-v1.json": "be1aac61c19ec6fe5adfcb8ce5fb18ba37820bd3e61fd72555f4ded69d04ac14",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/plan.json": "0c958315767a17a81f6af0c0a07bc1d8d592a7e65b2368b97d32de98f34b7777",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/records.jsonl": "40c3182a5069916cc22d4bf4383ed36cc8df8cbdd066a801ffa6765c149a0d6e",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/raw-audit-v1.json": "4cd22c0997ec75c3c7485c5ff7060ffbb7047d768c0d5a2032b7ea698a5d684b",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/observations-audit-v1.json": "af9824775d002e20dd8d2b4a8287d3a5c8ab6d647457c05ab2115b12a34471b4",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/report-diagnostic-v1.json": "42c66e207a9ced613fef9aa508decde5a9e8bfac2c6183bba09d67863080fa14",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation09-20260928/exit-projects-clear-v1.json": "8446850233129f00ae4656da0b93915f58d4c7cd0a9affc3d5621ee9226938d4",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/plan.json": "a70dd956b1302b482517f01d54665ac56f582ecca2b19ba905097f3efd29726a",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/records.jsonl": "d043fadf27c130e7174d0136ef2fa553820bb3c57f3c0886513bda481f0697ae",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/raw-audit-v1.json": "0abf6f6c84f3861bf84ef6b9230be5a016ef5ee8a44d6985d8cbe7246ad36b8c",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/observations-audit-v1.json": "554f7f4ad0621a4d4e339486a4772c29c9f2bbb2270b484737828e28f4a96094",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/report-diagnostic-v1.json": "056f5277239d76e2eaecae4ba9a3ca0e8980385f4433f69129b5d0889d7cff96",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation10-20260928/exit-projects-clear-v1.json": "d04e3aafe997e8f3f33acaa2f816466d7977b77376bc930f8dd031ce711484d1",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/plan.json": "20ec48944f0fbc4b8873b952fafcdb578326a91026d1da2c4ae8ff0c9de4c1d0",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/records.jsonl": "ddcfad2046e8ed5e2f526be8b4267725e0382870e1fc6d1eb375221e48f8b3c2",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/raw-audit-v1.json": "13b2090747240fd4bd58f9446e9dcfa2e3aefc2daa27034c20d84d116335f011",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/observations-audit-v1.json": "b18862cffc1a52d6371df0ced0ea3ff5f790cbfefcab3f31a71be11e344ba002",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/report-diagnostic-v1.json": "5971dc75792f412bb927217ca80cd517d09a6ea078c5b87c5d08834198b45ed5",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation11-20260928/exit-projects-clear-v1.json": "d3ac6ef95a80c61ca41fd3e76b5672a9f773b313cc00cbf5ece64c97af710b49",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/plan.json": "f0858fc8c5560d3a926b10c60095e50821deb29a5976dce4a8fea513e11318aa",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/records.jsonl": "8687dfcedccaa95e595cd787365e6da8356da933cd1f4bd1a7c70b59bb959be6",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/raw-audit-v1.json": "d7108eeebf42fc416c09011abacdf867c6c63e990af988a843913e1ecc0297f6",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/observations-audit-v1.json": "fb135937bcaf4cbd986f9f764dec5ebd0c514f005b5acf230ee7a6718617b40a",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/report-diagnostic-v1.json": "919ecbf0e9677009769d1d15cc6fa0624b701021195195a46f0d258c086ddb73",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation12-20260928/exit-projects-clear-v1.json": "a9d00b9f5c0dc69341e34ae01ede70e0173d8413a4600ca3688869488345ce3b",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/plan.json": "a607302a57cf8fa53313b501d05051f700ccb7b8f22d455dba2815a797dc0075",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/records.jsonl": "0cef894476f4b02ca3a72deef51dde7ff07a80c00f77f16d0c7703d8a3bac9c2",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/raw-audit-v1.json": "8932d6cf58b6f76a3fade822b94717d0609aec528570c1f2ba8433d2218ba2cb",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/observations-audit-v1.json": "9c13179b8334f88b5e0e7a3961c60ef18652091567edbbf7d66526cbec6e7be7",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/report-diagnostic-v1.json": "ef0e19f9961c46fe1602985d42a077bd4117f255e01617ab3f9cd18135c9d5e3",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation13-20260928/exit-projects-clear-v1.json": "a5ca24cd32c1e537af481e7547e456cfde752070cfb9a14d40d8366d0db900fc",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/plan.json": "3d9affd0e2d6f1a6c72b81e2543262a3280d31c034cd3d18f641fbe5005311e0",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/records.jsonl": "7f008c9033f94f6ff8c9d48b52f88b48e4d074a8088d3b9dee8d1476de53d317",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/raw-audit-v1.json": "6cf0e4ef5a57d85679161d349f836d4ec35f7b23d541d4564a08471a69a864d7",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/observations-audit-v1.json": "316e7186f43042ee05db3c8aa9b0620210d75a6ac6a097a98e2c579a5ec5b740",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/report-diagnostic-v1.json": "7baf3b595675232456ba45bd1624f1091e3f1a60af56de22c353ea39e3b59684",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation14-20260928/exit-projects-clear-v1.json": "d5039697cc157a738a094da8efbdf99821634f70196115e710f5ab9b864f59ba",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/plan.json": "a37dc21c31d5c1baa83e676f741c0664655018b34086accbf3fa3cbaf6ec0952",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/records.jsonl": "ff8a72f401cd988666d98ef51156657ecc9bcd0d5f22ad0efd5eabce2b4df8b0",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/raw-audit-v1.json": "b8ee853720ea19bb2415fb821d18374ae24f8c908c4adfc7680719af4ce1078b",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/observations-audit-v1.json": "c801000c85a9d22a5b2122f65baea6407d6014e72a19815f8f369b9f38350efc",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/report-diagnostic-v1.json": "4053957469a192ad7a43f75eebad85c6950b6a6343a8dcb28ac241c598d4aed7",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation15-20260928/exit-projects-clear-v1.json": "12e51f498d72e6f45c9eb42d575cd0ce32b36b12c59320a6831364b3d4253191",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/plan.json": "ec7d6403dbb7dd8e1f00f797bb5ccf01d0589aefff8b6196f1aefbf2eee621aa",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/records.jsonl": "6ddd026657151bb173e7ee2d41b77388f68ef97b3e34ee4bbf3f9036e278d240",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/raw-audit-v1.json": "d74f6e910bee854a53f39a0f5c014d8019f94116dc93f72023e514917b574d33",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/observations-audit-v1.json": "0576d41361dcb10fa1803be7e123c9d627dfa3886eab7fc0f1bb77b013f576ac",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/report-diagnostic-v1.json": "9184a8622461ce42ddb6dd16a71eb38e18c049907435b0b8321442bb340ee8ce",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation16-retry1-20260928/exit-projects-clear-v1.json": "873764d485831d9562eeb72abf5ebd0fb5fbd3c12810baa750ea090d32e07360",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/plan.json": "6db65d5c4476fa9d7188877448860ad4799a31800d4624e7da733a23f987dff3",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/records.jsonl": "fb15d1a5749d2b7f9c62e6d38b39a82cbfe11531846bbcb8931e91879ed3788f",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/raw-audit-v1.json": "273acf41c3ea62ee003694d574912d5986eecd4dd8f337caf2dc04c9a5bb0b92",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/observations-audit-v1.json": "59a940440722ad63c2958d172507723d4a34b0f15172c891071ca83129b27211",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/report-diagnostic-v1.json": "648ac2da24892f07d5cbf949ba8852575cbcf0b907da0ba5497ade75c41735c9",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation17-20260930/exit-projects-clear-v1.json": "5e6291676539ceff8cdf6ddd6601a0b2b08c724517e297e841905be432565e77",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/plan.json": "b5344862fd8c195f8f71b8b0948cfcc7daf1ddbf0390c27cd3bb6d19e2346468",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/records.jsonl": "6e87f0ece4b1eb6576c3f6016f3b85779348690064f69b02bf9d13412cf5bd58",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/raw-audit-v1.json": "3883953bcb8747dd02c0c84a78093fcb0701a553b41f2f144a8b5f5c10314900",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/observations-audit-v1.json": "ef8966fd843eb984c2a8825e69bf6bd1e89e72e7cdd99d837a4289f63e14468d",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/report-diagnostic-v1.json": "525ce850f5f0f5883d16661a71a4ec0aa1cd3d4d8efe83acbddd027c78e802b5",
    "/data1/home/zly/swe/rl/runs/rl-paper-20260927/fixtures/eval500-common-validation18-20260930/exit-projects-clear-v1.json": "a9f5c108f364d23503a62589faa4c3aef32364a99d0c303308f959df1fc459c6"
  },
  "metric_notes": {
    "strict_pass": "Complete PASSED/FAILED fixed observations AND expected gold1/empty0; excludes the fully observed Django10097 empty1",
    "report_binary": "Binary report diagnosis includes unexpected control scores; not a count of passed controls or formal model slots"
  },
  "aggregation_audit": {
    "method": "sum retained cohort summaries in cohorts[]; no imputation of unresolved or partial observations",
    "cohort_count": 23,
    "task_sum": 500,
    "control_sum": 1000,
    "previous_counts_corrected": {
      "controls": 1000,
      "raw_pass": 999,
      "strict_pass": 779,
      "report_binary": 798,
      "unresolved": 101,
      "partial_zeros": 18,
      "control_expectation_mismatches": 5,
      "binary_decisions_matching_control_expectation": 797
    }
  }
}

公开内容 SHA-256 45ef82e6cc1980428d679e28ad6f29ab4ec72fa843521448f25a5172a2385a48

返回资料与接手