{
  "model": "claude-opus-4-6",
  "api_base": "https://hche3637.com/v1",
  "timestamp": "2026-04-30T08:38:33.601696",
  "total_evaluated": 1200,
  "avg_latency_seconds": 25.45,
  "api_errors": 15,
  "summary": {
    "total_episodes": 1200,
    "detection_accuracy": 0.48583333333333334,
    "miss_rate": 0.06727272727272728,
    "false_alarm_rate": 0.8753846153846154,
    "precision": 0.47412199630314233,
    "recall": 0.9327272727272727,
    "f1_security": 0.6286764705882353,
    "threat_type_accuracy": 0.25146198830409355,
    "parse_failure_rate": 0.029166666666666667,
    "breakdown": {
      "TP": {
        "total": 550,
        "correct": 513,
        "accuracy": 0.9327272727272727
      },
      "FP": {
        "total": 400,
        "correct": 37,
        "accuracy": 0.0925
      },
      "TN": {
        "total": 250,
        "correct": 33,
        "accuracy": 0.132
      }
    },
    "per_sq": {
      "SQ1": {
        "total": 190,
        "accuracy": 0.45789473684210524,
        "miss_count": 17,
        "false_alarm_count": 81
      },
      "SQ2": {
        "total": 240,
        "accuracy": 0.4625,
        "miss_count": 5,
        "false_alarm_count": 124
      },
      "SQ3": {
        "total": 290,
        "accuracy": 0.46206896551724136,
        "miss_count": 7,
        "false_alarm_count": 148
      },
      "SQ4": {
        "total": 290,
        "accuracy": 0.4827586206896552,
        "miss_count": 3,
        "false_alarm_count": 145
      },
      "SQ5": {
        "total": 190,
        "accuracy": 0.5842105263157895,
        "miss_count": 5,
        "false_alarm_count": 71
      }
    }
  },
  "errors": {
    "total_errors": 1001,
    "error_distribution": {
      "FALSE_ALARM": 567,
      "PARSE_FAIL": 21,
      "WRONG_TYPE": 384,
      "MISS": 29
    },
    "error_by_sq": {
      "SQ1": {
        "FALSE_ALARM": 81,
        "PARSE_FAIL": 5,
        "WRONG_TYPE": 63,
        "MISS": 17
      },
      "SQ2": {
        "FALSE_ALARM": 124,
        "WRONG_TYPE": 57,
        "PARSE_FAIL": 3,
        "MISS": 2
      },
      "SQ3": {
        "FALSE_ALARM": 148,
        "PARSE_FAIL": 5,
        "WRONG_TYPE": 107,
        "MISS": 3
      },
      "SQ4": {
        "FALSE_ALARM": 143,
        "PARSE_FAIL": 4,
        "WRONG_TYPE": 112,
        "MISS": 3
      },
      "SQ5": {
        "FALSE_ALARM": 71,
        "PARSE_FAIL": 4,
        "WRONG_TYPE": 45,
        "MISS": 4
      }
    },
    "error_by_category": {
      "device_fault": 139,
      "fire_gas": 106,
      "intrusion": 86,
      "behavioral_anomaly": 38,
      "water_damage": 2,
      "child_specific": 15,
      "elderly_specific": 27
    }
  }
}