{
  "id": 4611,
  "url": "https://arxiv.org/abs/2605.09661v1",
  "title": "MedMeta: A Benchmark for LLMs in Synthesizing Meta-Analysis Conclusion from Medical Studies",
  "summary": "Large language models (LLMs) have saturated standard medical benchmarks that test factual recall, yet their ability to perform higher-order reasoning, such as synthesizing evidence from multiple sources, remains critically under-explored. To address this gap, we introduce MedMeta, the first benchmark designed to evaluate an LLM's ability to generate conclusions from medical meta-analyses using only the abstracts of cited studies. MedMeta comprises 81 meta-analyses from PubMed (2018--2025) and ev",
  "authors": "Huy Hoang Ha, Benoit Favre, Francois Portet",
  "category": "research",
  "topics": "healthcare",
  "orgs": "meta",
  "regions": null,
  "published_at": "2026-05-10T17:20:39.000Z",
  "fetched_at": "2026-07-14T16:31:08.355Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/4611",
  "original_url": "https://arxiv.org/abs/2605.09661v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}