{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-09-27-002",
    "source_story_id": "tmp-feature-enigmaforge-hidden-question",
    "edition_id": "mp-2026-09-27-morning-0080",
    "edition_url": "https://themachinepress.com/edition/2026-09-27",
    "position": 2,
    "story_type": "secondary",
    "section": "benchmarks-evals",
    "editorial_classification": "editorial",
    "headline": "The Benchmark Hid the Question",
    "slug": "the-benchmark-hid-the-question",
    "dek": "EnigmaForge asked models to infer both the puzzle and its answer from documents, producing a 22-fold spread in intuition.",
    "summary": "EnigmaForge asked models to infer both the puzzle and its answer from documents, producing a 22-fold spread in intuition.",
    "body_text": "EnigmaForge replaces the usual explicit question with a packet of documents whose facts imply both what must be solved and the one admissible answer. Its generator uses satisfiability checks to create puzzles with a unique solution and to verify that every clue is load-bearing. Across 25 frontier models, 600 instances and 17,400 evaluation records, the authors report a 22-fold spread in reconstructing the hidden problem, compared with a 1.6-fold spread in recovering stated facts. They also filter refusals separately so a model that declines the task is not confused with one that reasons incorrectly. This is a synthetic benchmark preprint, but it isolates a practical failure mode: gathering facts can be much easier than discovering the question those facts were meant to answer.",
    "why_it_matters": "EnigmaForge asked models to infer both the puzzle and its answer from documents, producing a 22-fold spread in intuition.",
    "limitations": [
      "They also filter refusals separately so a model that declines the task is not confused with one that reasons incorrectly."
    ],
    "importance": 10,
    "canonical_url": "https://themachinepress.com/story/mp-2026-09-27-002/the-benchmark-hid-the-question",
    "json_url": "https://themachinepress.com/story/mp-2026-09-27-002.json",
    "first_published_at": "2026-09-27T09:00:00.000-04:00",
    "modified_at": "2026-09-27T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-09-27-002-001",
        "text": "EnigmaForge asked models to infer both the puzzle and its answer from documents, producing a 22-fold spread in intuition.",
        "source_ids": [
          "source-2026-09-27-002"
        ],
        "qualification": "They also filter refusals separately so a model that declines the task is not confused with one that reasons incorrectly."
      }
    ],
    "source_ids": [
      "source-2026-09-27-002"
    ],
    "tags": [
      "reasoning benchmarks",
      "latent questions",
      "document understanding"
    ],
    "image_url": "https://themachinepress.com/issues/2026-09-27/feature-enigmaforge-hidden-question.png",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-09-27-002",
      "title": "arXiv preprint 2609.30144",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2609.30144",
      "canonical_url": "https://arxiv.org/abs/2609.30144",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-09-24T13:10:03.000-04:00",
      "accessed_at": "2026-09-27T08:26:41.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-09-27-002-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Benchmark Hid the Question",
    "publisher": "The Machine Press",
    "published_at": "2026-09-27T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-09-27-002/the-benchmark-hid-the-question"
  }
}