{
  "model": "Qwen/Qwen3.6-35B-A3B-FP8",
  "api_base": "http://localhost:8000/v1",
  "timestamp": "2026-05-05T02:58:52.375174",
  "total_evaluated": 60,
  "summary": {
    "total_episodes": 60,
    "detection_accuracy": 0.6166666666666667,
    "miss_rate": 0.39285714285714285,
    "false_alarm_rate": 0.375,
    "precision": 0.5862068965517241,
    "recall": 0.6071428571428571,
    "f1_security": 0.5964912280701754,
    "threat_type_accuracy": 0.4117647058823529,
    "parse_failure_rate": 0.0,
    "breakdown": {
      "TP": {
        "total": 28,
        "correct": 17,
        "accuracy": 0.6071428571428571
      },
      "FP": {
        "total": 28,
        "correct": 18,
        "accuracy": 0.6428571428571429
      },
      "TN": {
        "total": 4,
        "correct": 2,
        "accuracy": 0.5
      }
    },
    "per_sq": {
      "SQ1": {
        "total": 13,
        "accuracy": 0.6153846153846154,
        "miss_count": 5,
        "false_alarm_count": 0
      },
      "SQ2": {
        "total": 6,
        "accuracy": 0.6666666666666666,
        "miss_count": 1,
        "false_alarm_count": 1
      },
      "SQ3": {
        "total": 17,
        "accuracy": 0.5294117647058824,
        "miss_count": 3,
        "false_alarm_count": 5
      },
      "SQ4": {
        "total": 17,
        "accuracy": 0.5882352941176471,
        "miss_count": 2,
        "false_alarm_count": 5
      },
      "SQ5": {
        "total": 7,
        "accuracy": 0.8571428571428571,
        "miss_count": 0,
        "false_alarm_count": 1
      }
    }
  },
  "errors": {
    "total_errors": 33,
    "error_distribution": {
      "MISS": 11,
      "FALSE_ALARM": 12,
      "WRONG_TYPE": 10
    },
    "error_by_sq": {
      "SQ1": {
        "MISS": 5
      },
      "SQ2": {
        "FALSE_ALARM": 1,
        "MISS": 1,
        "WRONG_TYPE": 1
      },
      "SQ3": {
        "FALSE_ALARM": 5,
        "MISS": 3,
        "WRONG_TYPE": 3
      },
      "SQ4": {
        "FALSE_ALARM": 5,
        "MISS": 2,
        "WRONG_TYPE": 4
      },
      "SQ5": {
        "FALSE_ALARM": 1,
        "WRONG_TYPE": 2
      }
    },
    "error_by_category": {
      "device_fault": 5,
      "intrusion": 4,
      "water_damage": 1,
      "behavioral_anomaly": 3,
      "child_specific": 4,
      "elderly_specific": 3,
      "fire_gas": 1
    }
  },
  "pipeline": "EGPv2.1"
}