{
  "title": "Android SO：保护前后实测对照",
  "model": "gpt-5.6-sol · high",
  "before_label": "未保护",
  "after_label": "虚拟化保护",
  "source": "同秘密、同轨道配对；18 次方案对照 + 6 次未保护基线；2026-10-30",
  "rows": [
    {
      "instance": 0,
      "track": "S_static",
      "before": {
        "trial_id": "34ec6d3176cc4c07",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 44.98233829200035,
        "tool_calls": 4,
        "usage": {
          "input_tokens": 52999,
          "cached_input_tokens": 40320,
          "cache_write_input_tokens": 0,
          "output_tokens": 1725,
          "reasoning_output_tokens": 1003
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "dffa45c5660aa227",
        "correct": 6057,
        "total": 8592,
        "elapsed_seconds": 496.7499370419828,
        "tool_calls": 48,
        "usage": {
          "input_tokens": 2513332,
          "cached_input_tokens": 2401792,
          "cache_write_input_tokens": 0,
          "output_tokens": 32172,
          "reasoning_output_tokens": 17901
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": false,
        "invalid_reason": "静态轨实际执行目标 VM 程序，排除比较",
        "outcome": "CONTAMINATED"
      }
    },
    {
      "instance": 0,
      "track": "D_dynamic",
      "before": {
        "trial_id": "70c9b9dc98775ec2",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 67.24974116700469,
        "tool_calls": 8,
        "usage": {
          "input_tokens": 120732,
          "cached_input_tokens": 108160,
          "cache_write_input_tokens": 0,
          "output_tokens": 3218,
          "reasoning_output_tokens": 920
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "f9d0696a64d0db5c",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 550.4473012909875,
        "tool_calls": 58,
        "usage": {
          "input_tokens": 5281944,
          "cached_input_tokens": 5118976,
          "cache_write_input_tokens": 0,
          "output_tokens": 27410,
          "reasoning_output_tokens": 11877
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      }
    },
    {
      "instance": 1,
      "track": "S_static",
      "before": {
        "trial_id": "68d02c81a7d65983",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 36.081048458989244,
        "tool_calls": 4,
        "usage": {
          "input_tokens": 52860,
          "cached_input_tokens": 35968,
          "cache_write_input_tokens": 0,
          "output_tokens": 1484,
          "reasoning_output_tokens": 797
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "595383454349b321",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 678.8936557919951,
        "tool_calls": 57,
        "usage": {
          "input_tokens": 4413848,
          "cached_input_tokens": 4268544,
          "cache_write_input_tokens": 0,
          "output_tokens": 50224,
          "reasoning_output_tokens": 19916
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      }
    },
    {
      "instance": 1,
      "track": "D_dynamic",
      "before": {
        "trial_id": "7f93b9f3cfc2fe82",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 53.59830199999851,
        "tool_calls": 4,
        "usage": {
          "input_tokens": 53122,
          "cached_input_tokens": 41728,
          "cache_write_input_tokens": 0,
          "output_tokens": 2004,
          "reasoning_output_tokens": 1303
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "8674d65e676be161",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 443.203093041986,
        "tool_calls": 40,
        "usage": {
          "input_tokens": 2712135,
          "cached_input_tokens": 2569472,
          "cache_write_input_tokens": 0,
          "output_tokens": 25227,
          "reasoning_output_tokens": 9884
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      }
    },
    {
      "instance": 2,
      "track": "S_static",
      "before": {
        "trial_id": "d2da649c50a202ba",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 39.76116570801241,
        "tool_calls": 5,
        "usage": {
          "input_tokens": 72543,
          "cached_input_tokens": 59008,
          "cache_write_input_tokens": 0,
          "output_tokens": 2051,
          "reasoning_output_tokens": 1296
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "4460daf851bb5d05",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 487.197847250005,
        "tool_calls": 33,
        "usage": {
          "input_tokens": 2201094,
          "cached_input_tokens": 2083968,
          "cache_write_input_tokens": 0,
          "output_tokens": 33467,
          "reasoning_output_tokens": 19513
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      }
    },
    {
      "instance": 2,
      "track": "D_dynamic",
      "before": {
        "trial_id": "68c7c2d520d448e8",
        "correct": 8592,
        "total": 8592,
        "elapsed_seconds": 30.114392749994295,
        "tool_calls": 4,
        "usage": {
          "input_tokens": 52777,
          "cached_input_tokens": 41472,
          "cache_write_input_tokens": 0,
          "output_tokens": 1470,
          "reasoning_output_tokens": 892
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "VERIFIED_RECOVERY"
      },
      "after": {
        "trial_id": "20a77cacf48a0bd8",
        "correct": 5697,
        "total": 8592,
        "elapsed_seconds": 330.17099312497885,
        "tool_calls": 25,
        "usage": {
          "input_tokens": 669419,
          "cached_input_tokens": 624768,
          "cache_write_input_tokens": 0,
          "output_tokens": 12330,
          "reasoning_output_tokens": 4135
        },
        "runtime_equivalence": {
          "correct": 8592,
          "total": 8592
        },
        "eligible": true,
        "invalid_reason": "",
        "outcome": "WRONG_CONFIDENT"
      }
    }
  ]
}
