agent.run

{
  "feature": "search_qa",
  "user_id": "u_heavy_18"
}

{
  "model": "gpt-4o",
  "feature": "search_qa",
  "user_id": "u_heavy_18",
  "tokens_prompt": 357,
  "tokens_output": 478,
  "tokens_total": 835,
  "input": "Summarize the Q3 earnings report attached. Focus on revenue, margin, and guidance.",
  "output": "(answer to: Summarize the Q3 earnings report attache…)",
  "eval_scores": {
    "Hallucination": 0.7755703721195459
  },
  "hallucination_details": {
    "claims": [
      {
        "text": "Day 1 plan fits under the $200/day soft cap.",
        "verdict": "supported"
      },
      {
        "text": "Belém Tower opens at 9am.",
        "verdict": "contradicted"
      },
      {
        "text": "Time Out Market lunch costs $18.",
        "verdict": "unsupported"
      },
      {
        "text": "Castelo dinner costs $42.",
        "verdict": "unsupported"
      }
    ],
    "supported": 1,
    "contradicted": 1,
    "unsupported": 2,
    "total": 4,
    "score": 0.25
  }
}

{
  "model": "gpt-4o",
  "feature": "search_qa",
  "user_id": "u_heavy_18",
  "tokens_prompt": 631,
  "tokens_output": 20,
  "tokens_total": 651,
  "input": "Given the customer ticket below, draft a refund response that follows policy P-204.",
  "output": "(answer to: Given the customer ticket below, draft a…)",
  "eval_scores": {
    "Hallucination": 0.5379494426771998
  }
}