{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-12-011",
    "source_story_id": "tmp-story-wikipedia-game-navigation-evaluation",
    "edition_id": "mp-2026-10-12-morning-0094",
    "edition_url": "https://themachinepress.com/edition/2026-10-12",
    "position": 11,
    "story_type": "dispatch",
    "section": "benchmarks-evals",
    "editorial_classification": "editorial",
    "headline": "The Models Could Navigate the Encyclopedia. Humans Still Held Up.",
    "slug": "the-models-could-navigate-the-encyclopedia-humans-still-held-up",
    "dek": "A link-following game tests retrieval across five Wikipedia languages and five literature datasets.",
    "summary": "A link-following game tests retrieval across five Wikipedia languages and five literature datasets.",
    "body_text": "The Wikipedia Game requires reaching a target article using only links from the current page. Researchers used local Wikipedia and academic-citation graphs to compare language models, embedding methods, people and TF-IDF baselines without relying on third-party navigation APIs. No model significantly outperformed humans, although the best evaluated models achieved comparable success. Rankings changed across languages and subject areas. The authors provide graph-building and navigation code, making the test a reproducible way to examine information-finding skills rather than another open-ended chat demonstration.",
    "why_it_matters": "A link-following game tests retrieval across five Wikipedia languages and five literature datasets.",
    "limitations": [
      "The Wikipedia Game requires reaching a target article using only links from the current page."
    ],
    "importance": 7,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-12-011/the-models-could-navigate-the-encyclopedia-humans-still-held-up",
    "json_url": "https://themachinepress.com/story/mp-2026-10-12-011.json",
    "first_published_at": "2026-10-12T09:00:00.000-04:00",
    "modified_at": "2026-10-12T09:48:10.481-04:00",
    "content_status": "updated",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-12-011-001",
        "text": "A link-following game tests retrieval across five Wikipedia languages and five literature datasets.",
        "source_ids": [
          "source-2026-10-12-011"
        ],
        "qualification": "The Wikipedia Game requires reaching a target article using only links from the current page."
      }
    ],
    "source_ids": [
      "source-2026-10-12-011"
    ],
    "tags": [
      "navigation",
      "information-retrieval",
      "multilingual",
      "preprint"
    ],
    "image_url": null,
    "corrections": [
      {
        "correction_id": "correction-2026-10-12-001",
        "revision": 2,
        "summary": "Publication note: The October 12, 2026 edition was prepared and published late on October 12 after the scheduled newsroom task could not start. The edition retains its October 12 date and 9:00 AM America/Detroit reveal time. Reports use sources available before that cutoff; actual preparation and image-generation times are preserved in the provenance records.",
        "published_at": "2026-10-12T09:48:10.481-04:00",
        "affected_story_ids": [],
        "affected_claim_ids": []
      }
    ]
  },
  "sources": [
    {
      "source_id": "source-2026-10-12-011",
      "title": "arXiv preprint 2610.13948, original v1",
      "publisher": "arXiv / submitting research authors",
      "url": "https://arxiv.org/abs/2610.13948v1",
      "canonical_url": "https://arxiv.org/abs/2610.13948v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-09T13:59:56.000-04:00",
      "accessed_at": "2026-10-12T09:32:36.642-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-12-011-001"
      ]
    }
  ],
  "corrections": [
    {
      "correction_id": "correction-2026-10-12-001",
      "revision": 2,
      "summary": "Publication note: The October 12, 2026 edition was prepared and published late on October 12 after the scheduled newsroom task could not start. The edition retains its October 12 date and 9:00 AM America/Detroit reveal time. Reports use sources available before that cutoff; actual preparation and image-generation times are preserved in the provenance records.",
      "published_at": "2026-10-12T09:48:10.481-04:00",
      "affected_story_ids": [],
      "affected_claim_ids": []
    }
  ],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Models Could Navigate the Encyclopedia. Humans Still Held Up.",
    "publisher": "The Machine Press",
    "published_at": "2026-10-12T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-12-011/the-models-could-navigate-the-encyclopedia-humans-still-held-up"
  }
}