{
  "attempts": 809,
  "cached_cost_usd": null,
  "corpus": {
    "name": "RouterArena · sub_10 integrado",
    "revision": "cff9659dd6a3a07815f4842fcdadea5a597a1df1",
    "selection": "Todos los 809 IDs de Claude, iguales a Gemini; GPT mini se filtra a esos IDs; orden de Claude; igualdad exacta de preguntas",
    "source_url": "https://github.com/RouteWorks/RouterArena/tree/cff9659dd6a3a07815f4842fcdadea5a597a1df1",
    "total": 809
  },
  "decisions": 809,
  "distribution": [
    {
      "count": 809,
      "model_id": "claude-3-haiku-20240307",
      "target_model": "claude-3-haiku-20240307"
    },
    {
      "count": 0,
      "model_id": "gemini-2.0-flash-001",
      "target_model": "gemini-2.0-flash-001"
    },
    {
      "count": 0,
      "model_id": "gpt-4o-mini",
      "target_model": "gpt-4o-mini"
    }
  ],
  "errors": 1,
  "evaluation_completed": false,
  "failure": {
    "action": "complete_evaluation_failed; no_clipping_no_exclusion_no_imputation",
    "affected_model": "claude-3-haiku-20240307",
    "cached_score": -0.0330188679245283,
    "code": "failed_preregistered_quality_domain",
    "diagnostic_sha256": "af57d9a2d0edc761d34a44c79862bf7b72bb01855d7ec97390055d81e4e9d927",
    "expected_domain": [
      0,
      1
    ],
    "global_index": "NarrativeQA_4102",
    "invalid_measurements": 1,
    "measurement_denominator": 2427,
    "metric": "meteor_score"
  },
  "generated_at": "2026-09-25T12:03:38.183907+00:00",
  "inferences": 0,
  "integrity": {
    "allowed_model_ids": [
      "claude-3-haiku-20240307",
      "gemini-2.0-flash-001",
      "gpt-4o-mini"
    ],
    "availability_assumption": "reference_membership_only; no_current_availability_claim",
    "capability_gates_applied": false,
    "frozen_hour": 10,
    "guards": {
      "execution_attempts": 0,
      "network_attempts": 0,
      "store_access_attempts": 0,
      "write_attempts": 0
    },
    "identity_injection": "production_default_without_user_profile",
    "learning_mode_preserved": "on",
    "network_disabled": true,
    "no_labels_or_outcomes": true,
    "objective": "balanced",
    "offline_only": true,
    "output_limit": null,
    "plan_id": "pro-code",
    "reference_arm_matching": "exact_id_only; no_family_mapping",
    "reference_pool_sha256": "58487ec54880fb2a2b5391f348130c20756cc37262113d2bee830bbf28d53e3f",
    "selector_scope": "historical_reference_pool_pure_formula",
    "snapshot_sha256": "dd7ce8026b1a8b31a9edd339a7c98f046e7b85779b6103956fbb68b3297aa3a8",
    "source_sha256": {
      "theus_oauth_server.py": "fb318a5f8c4bacfec40e15b0057c49fc01e8e9245924ee6ffc1d19f7ca8e0a13",
      "titan_contract.py": "6499a5a8bd102eec530ff5c93fbf5e47419f81bd150a58b4c02867c7b30c9004",
      "titan_evidence.py": "997d5deb99cf43ea8d2ef29e7477fe50f8c61cee8c5196aacc9f82457877b742",
      "titan_evidence_http.py": "814b3b53b8357c8cc0375d0d83d7c2e07374b9ab83b259f285b37df8f63e4fad",
      "titan_gateway_adapter.py": "b13dc2e80f29bb8f15d3e8999df2714926adad0b9fd8b837a83804e34fa97c74"
    },
    "state_sha256_after": "e731b6ae2085db58b5f71e338a4a51c222a75658918ae0e7611baec930ae173c",
    "state_sha256_before": "e731b6ae2085db58b5f71e338a4a51c222a75658918ae0e7611baec930ae173c",
    "strategy": "single",
    "training_disabled": true,
    "unchanged": true,
    "writes_disabled": true
  },
  "latency": {
    "median_ms": 0.542208000000155,
    "p95_ms": 1.105875000000145,
    "samples": 809,
    "scope": "local_selector_only"
  },
  "limitations": [
    "Subconjunto integrado sub_10 de 809 casos; no es score oficial completo ni ranking de RouterArena",
    "Los tres arms son frescos; sin entrenamiento o ajuste con el conjunto",
    "Costes y calidad corresponden a respuestas históricas cacheadas, no a inferencias de TITAN",
    "Latencia mide sólo el selector local",
    "Evaluación detenida según prerregistro: un score de la caché oficial viola el dominio [0,1]. No se publica ninguna media ni comparación de calidad."
  ],
  "methodology": [
    "Sólo question: sin IDs, etiquetas, respuestas, costes por caso, eval_name ni resultados",
    "Media aritmética de evaluation_result.score cacheado en [0,1], sin transformar; correct=null; no se presupone clasificación binaria",
    "Suma de evaluation_result.inference_cost cacheado en USD; cero gasto nuevo; TT sólo es peso interno, no precio USD",
    "Cualquier ausencia, duplicado, pregunta diferente, coste inválido o score inválido hace fallar la ejecución completa"
  ],
  "new_spend_usd": 0,
  "predictions_seal_sha256": "ab32d275e4680a38dbf3e374e48a7ea9185d8e5749ab8e17f66c30c4b6e5f9b3",
  "protocol_sha256": "9475e52b838ae6dd7c5dbb69ab9134f8de75adcc18503c12676828e891a5e078",
  "quality_score": null,
  "results": [],
  "schema_version": 1,
  "scope": "internal_offline_historical_reference_not_official_arena_score",
  "status": "failed_preregistered_quality_domain"
}
