{
 "at": 1794625193.9427366,
 "calibration_table": {
  "qwen2.5-0.5b": {
   "arithmetic": {
    "accuracy": 0.5,
    "correct": 1,
    "mean_latency_s": 189.2612,
    "tokens_out": 6,
    "total": 2
   },
   "classification": {
    "accuracy": 1.0,
    "correct": 2,
    "mean_latency_s": 0.7249,
    "tokens_out": 4,
    "total": 2
   },
   "code": {
    "accuracy": 1.0,
    "correct": 1,
    "mean_latency_s": 1.8266,
    "tokens_out": 2,
    "total": 1
   },
   "extraction": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 18.0057,
    "tokens_out": 48,
    "total": 1
   },
   "instruction": {
    "accuracy": 1.0,
    "correct": 1,
    "mean_latency_s": 1.5294,
    "tokens_out": 2,
    "total": 1
   },
   "reasoning": {
    "accuracy": 1.0,
    "correct": 1,
    "mean_latency_s": 1.7786,
    "tokens_out": 2,
    "total": 1
   }
  },
  "smollm2-135m": {
   "arithmetic": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 12.0796,
    "tokens_out": 63,
    "total": 2
   },
   "classification": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 0.7708,
    "tokens_out": 18,
    "total": 2
   },
   "code": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 2.7377,
    "tokens_out": 22,
    "total": 1
   },
   "extraction": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 8.0374,
    "tokens_out": 48,
    "total": 1
   },
   "instruction": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 1.3438,
    "tokens_out": 13,
    "total": 1
   },
   "reasoning": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 2.4102,
    "tokens_out": 19,
    "total": 1
   }
  },
  "smollm2-360m": {
   "arithmetic": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 13.2752,
    "tokens_out": 16,
    "total": 2
   },
   "classification": {
    "accuracy": 0.5,
    "correct": 1,
    "mean_latency_s": 0.6904,
    "tokens_out": 4,
    "total": 2
   },
   "code": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 3.2647,
    "tokens_out": 14,
    "total": 1
   },
   "extraction": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 1.9544,
    "tokens_out": 9,
    "total": 1
   },
   "instruction": {
    "accuracy": 0.0,
    "correct": 0,
    "mean_latency_s": 1.921,
    "tokens_out": 8,
    "total": 1
   },
   "reasoning": {
    "accuracy": 1.0,
    "correct": 1,
    "mean_latency_s": 0.6581,
    "tokens_out": 2,
    "total": 1
   }
  }
 },
 "calls": 8,
 "decision": "candidate policy; promotion requires external approval and a fresh test split",
 "delta_P2_vs_P0": 0.0,
 "delta_P2_vs_P1": 0.25,
 "execution_classes": {
  "REAL_LOCAL_MODEL": 8
 },
 "heldout_tasks": [
  "h1",
  "h2",
  "h3",
  "h4",
  "h5",
  "h6",
  "h7",
  "h8"
 ],
 "min_samples": 2,
 "note": "8 held-out prompts; small-sample result, reported with a Wilson interval, not a general claim",
 "policies": {
  "P0_fixed_best_model": {
   "accuracy": 0.5,
   "model": "qwen2.5-0.5b"
  },
  "P1_family_no_minimum": {
   "accuracy": 0.25
  },
  "P2_calibrated_min2": {
   "accuracy": 0.5,
   "rows": [
    {
     "basis": "family accuracy with >=2 samples",
     "correct": false,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "arithmetic",
     "latency_s": 16.2414,
     "model": "qwen2.5-0.5b",
     "task_id": "h1",
     "text": "92"
    },
    {
     "basis": "family accuracy with >=2 samples",
     "correct": false,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "arithmetic",
     "latency_s": 0.5849,
     "model": "qwen2.5-0.5b",
     "task_id": "h2",
     "text": "42"
    },
    {
     "basis": "family accuracy with >=2 samples",
     "correct": false,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "classification",
     "latency_s": 0.8638,
     "model": "qwen2.5-0.5b",
     "task_id": "h3",
     "text": "Anomalous Login"
    },
    {
     "basis": "family accuracy with >=2 samples",
     "correct": false,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "classification",
     "latency_s": 0.6398,
     "model": "qwen2.5-0.5b",
     "task_id": "h4",
     "text": "Security"
    },
    {
     "basis": "overall measured accuracy (family has too few samples)",
     "correct": true,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "extraction",
     "latency_s": 6.4394,
     "model": "qwen2.5-0.5b",
     "task_id": "h5",
     "text": "```json\n{\n  \"tools\": [\n    \"curl\",\n    \"awk\"\n  ]\n}\n```"
    },
    {
     "basis": "overall measured accuracy (family has too few samples)",
     "correct": true,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "reasoning",
     "latency_s": 2.1681,
     "model": "qwen2.5-0.5b",
     "task_id": "h6",
     "text": "False."
    },
    {
     "basis": "overall measured accuracy (family has too few samples)",
     "correct": true,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "instruction",
     "latency_s": 2.2054,
     "model": "qwen2.5-0.5b",
     "task_id": "h7",
     "text": "DONE"
    },
    {
     "basis": "overall measured accuracy (family has too few samples)",
     "correct": true,
     "execution_class": "REAL_LOCAL_MODEL",
     "family": "code",
     "latency_s": 2.3213,
     "model": "qwen2.5-0.5b",
     "task_id": "h8",
     "text": "4"
    }
   ],
   "wilson95": [
    0.2152,
    0.7848
   ]
  }
 },
 "schema": "cain72.router_policy.v1"
}