{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-26-012",
    "source_story_id": "tmp-story-rxnexbench-chemical-extraction",
    "edition_id": "mp-2026-10-26-morning-0108",
    "edition_url": "https://themachinepress.com/edition/2026-10-26",
    "position": 12,
    "story_type": "dispatch",
    "section": "research",
    "editorial_classification": "editorial",
    "headline": "Reading the Molecule Is Not Enough to Read the Reaction",
    "slug": "reading-the-molecule-is-not-enough-to-read-the-reaction",
    "dek": "RxnExBench separates molecular recognition from chemistry-aware reasoning across 3,565 annotated reactions and 26 structured fields.",
    "summary": "RxnExBench separates molecular recognition from chemistry-aware reasoning across 3,565 annotated reactions and 26 structured fields.",
    "body_text": "Synthetic-chemistry papers often compress conditions and outcomes into dense screen tables and substrate-scope figures. RxnExBench turns 3,565 expert-annotated reactions and 2,361 structures into a schema with 15 roles and 26 core fields. An oracle protocol supplies molecular recognition so the evaluation can isolate higher-level multimodal reasoning. Across nine vision-language models, performance fell as schema complexity and reasoning demands increased; even with oracle molecules, integrating the figure and applying chemistry knowledge remained the bottleneck. The benchmark is drawn from selected journals and annotations, and the paper promises public release; it does not show safe autonomous extraction for all chemistry literature.",
    "why_it_matters": "RxnExBench separates molecular recognition from chemistry-aware reasoning across 3,565 annotated reactions and 26 structured fields.",
    "limitations": [
      "The benchmark is drawn from selected journals and annotations, and the paper promises public release; it does not show safe autonomous extraction for all chemistry literature."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-26-012/reading-the-molecule-is-not-enough-to-read-the-reaction",
    "json_url": "https://themachinepress.com/story/mp-2026-10-26-012.json",
    "first_published_at": "2026-10-26T09:00:00.000-04:00",
    "modified_at": "2026-10-26T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-26-012-001",
        "text": "RxnExBench separates molecular recognition from chemistry-aware reasoning across 3,565 annotated reactions and 26 structured fields.",
        "source_ids": [
          "source-2026-10-26-012"
        ],
        "qualification": "The benchmark is drawn from selected journals and annotations, and the paper promises public release; it does not show safe autonomous extraction for all chemistry literature."
      }
    ],
    "source_ids": [
      "source-2026-10-26-012"
    ],
    "tags": [
      "chemistry",
      "vision-language models",
      "benchmarks",
      "scientific literature"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-26-012",
      "title": "arXiv preprint 2610.32221, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.32221v1",
      "canonical_url": "https://arxiv.org/abs/2610.32221v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-23T02:19:56.000-04:00",
      "accessed_at": "2026-10-26T08:26:57.002-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-26-012-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Reading the Molecule Is Not Enough to Read the Reaction",
    "publisher": "The Machine Press",
    "published_at": "2026-10-26T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-26-012/reading-the-molecule-is-not-enough-to-read-the-reaction"
  }
}