{
  "test": "quality-graded",
  "title": "Quality of finished work: Claims backed by source",
  "definition": "Share of graded claims backed by the source cited by the agent. A different model grades the claim without seeing the condition.",
  "denominator": "Graded claims: 536 without USDC; 559 with USDC. Limited to task types where both conditions had graded claims: market map, meeting prep, prospects and eligible Reddit runs.",
  "snapshot": "2026-11-13 16:59 UTC",
  "scope": "Seven models, 21 tasks, five repeats per condition; 1,470 runs in total. Completion and cost give tasks and models equal weight.",
  "availability": "These are the published run aggregates, not individual run logs. The source snapshot does not include per-run records, prompts, outputs or traces. No individual run records have been reconstructed.",
  "units": "Scores are fractions from 0 to 1. Changes ending in _pp are percentage points; counts are eligible runs or claims as defined above.",
  "source": {
    "study": "/experiments/funded-agent-index",
    "methodology": "/experiments/funded-agent-index#spend-top"
  },
  "chartUrl": "/experiments/funded-agent-index?quality=graded#quality",
  "downloads": {
    "json": "/experiments/funded-agent-index/data/quality-graded.json",
    "csv": "/experiments/funded-agent-index/data/quality-graded.csv"
  },
  "rows": [
    {
      "model": "Claude Opus 5.5",
      "model_id": "claude-opus-5-5",
      "without_usdc": 1,
      "with_usdc": 0.9333333333333333,
      "change_pp": -6.666666666666665,
      "without_eligible_count": 86,
      "with_eligible_count": 75
    },
    {
      "model": "Claude Sonnet 5.5",
      "model_id": "claude-sonnet-5-5",
      "without_usdc": 1,
      "with_usdc": 1,
      "change_pp": 0,
      "without_eligible_count": 80,
      "with_eligible_count": 86
    },
    {
      "model": "Claude Haiku 5.5",
      "model_id": "claude-haiku-5-5",
      "without_usdc": 1,
      "with_usdc": 0.9875,
      "change_pp": -1.2499999999999956,
      "without_eligible_count": 77,
      "with_eligible_count": 80
    },
    {
      "model": "Grok 4.6",
      "model_id": "grok-4.6",
      "without_usdc": 1,
      "with_usdc": 0.9878048780487805,
      "change_pp": -1.2195121951219523,
      "without_eligible_count": 71,
      "with_eligible_count": 82
    },
    {
      "model": "GPT-6 Luna",
      "model_id": "gpt-6-luna",
      "without_usdc": 0.9861111111111112,
      "with_usdc": 0.9722222222222222,
      "change_pp": -1.388888888888895,
      "without_eligible_count": 72,
      "with_eligible_count": 72
    },
    {
      "model": "GPT-6.1 Sol",
      "model_id": "gpt-6.1-sol",
      "without_usdc": 1,
      "with_usdc": 1,
      "change_pp": 0,
      "without_eligible_count": 73,
      "with_eligible_count": 86
    },
    {
      "model": "Gemini 3 Flash",
      "model_id": "gemini-3-flash-preview",
      "without_usdc": 0.948051948051948,
      "with_usdc": 0.9615384615384616,
      "change_pp": 1.348651348651353,
      "without_eligible_count": 77,
      "with_eligible_count": 78
    },
    {
      "model": "All models (pooled)",
      "model_id": "all",
      "without_usdc": 0.9906716417910447,
      "with_usdc": 0.9785330948121646,
      "change_pp": -1.2138546978880171,
      "without_eligible_count": 536,
      "with_eligible_count": 559
    }
  ],
  "taskRows": []
}