Step 10: tool results

search_similar_claims

{
  "limit": 6,
  "query": "METR time horizon doubling confidence intervals benchmark version uncertainty"
}

{
  "query": "METR time horizon doubling confidence intervals benchmark version uncertainty",
  "results": [
    {
      "id": "1d21c694-7f35-4518-ba59-770d0cfef986",
      "kind": "claim",
      "text": "METR Time Horizon 1.1's 50% time horizons above roughly 16 human-hours are unreliable measurements.",
      "claim_type": "empirical_verifiable",
      "state": "active",
      "similarity_score": 0.6932743288899795,
      "importance": 0.3,
      "assessment_status": null,
      "assessment_confidence": null
    },
    {
      "id": "d8be49b1-0f3c-43e5-8426-cc8bc25d9165",
      "kind": "quantity",
      "text": "Doubling time, in days, of the 50%-task-completion horizon of frontier AI models released from 2024 onward on METR's software task suite",
      "claim_type": "empirical_derived",
      "state": "active",
      "similarity_score": 0.5670859705285051,
      "importance": 0.55,
      "assessment_status": null,
      "assessment_confidence": null
    },
    {
      "id": "fab14deb-b0f5-45e4-bddc-a8c07d3a1719",
      "kind": "quantity",
      "text": "Doubling time of the 50%-task-completion time horizon of frontier AI models on METR's software task suite, 2019–2025, in days",
      "claim_type": "empirical_derived",
      "state": "active",
      "similarity_score": 0.5644276285706689,
      "importance": 0.65,
      "assessment_status": null,
      "assessment_confidence": null
    },
    {
      "id": "73fad761-cd72-4aaf-9484-39da8c3558c1",
      "kind": "claim",
      "text": "Dated frontier-AI forecast intervals often omit uncertainty from instrument, protocol, and data provenance choices.",
      "claim_type": "evaluative",
      "state": "active",
      "similarity_score": 0.558768625420639,
      "importance": 0.6,
      "assessment_status": null,
      "assessment_confidence": null
    },
    {
      "id": "434abbef-6a37-49af-acc9-7a534454b6b5",
      "kind": "claim",
      "text": "Benchmarks beyond METR's software suite show AI time horizons or capabilities growing faster since 2024.",
      "claim_type": "empirical_derived",
      "state": "active",
      "similarity_score": 0.5551512588524447,
      "importance": 0.3,
      "assessment_status": null,
      "assessment_confidence": null
    },
    {
      "id": "a78dc4eb-d2da-49a8-b15b-b525779ed539",
      "kind": "claim",
      "text": "Frontier AI 50% time horizons on METR's task suite have grown approximately exponentially with release date since 2024.",
      "claim_type": "empirical_derived",
      "state": "active",
      "similarity_score": 0.5339251101376392,
      "importance": 0.45,
      "assessment_status": "supported",
      "assessment_confidence": 0.65
    }
  ],
  "count": 6
}