{
  "id": 6095322,
  "title": "Molecular Déjà Vu: Digit-Level Retrieval of Published Values in Frontier Language Models",
  "url": "https://urgent.news/2026/09/04/molecular-deja-vu-digit-level-retrieval-of-published-values-in",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-04T17:32:48.000Z",
  "source": {
    "name": "arXiv cs.AI",
    "slug": "arxiv-cs-ai",
    "url": "https://arxiv.org/abs/2609.05381v1"
  },
  "original_language": "en",
  "account": null,
  "summary": "Large language models (LLMs) are increasingly evaluated on molecular property benchmarks, but accuracy cannot distinguish a model that predicts a property from one that retrieves a published number. We audit 22 frontier models on 12 regression benchmarks for verbatim retrieval and find that it is widespread but relatively benchmark-specific: on five datasets more than $50\\%$ of the LLMs show…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}