{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-21-008",
    "source_story_id": "tmp-story-foundpde-symbolic-laws",
    "edition_id": "mp-2026-10-21-morning-0103",
    "edition_url": "https://themachinepress.com/edition/2026-10-21",
    "position": 8,
    "story_type": "dispatch",
    "section": "research",
    "editorial_classification": "editorial",
    "headline": "A Foundation Model Writes the Equation Behind the Data",
    "slug": "a-foundation-model-writes-the-equation-behind-the-data",
    "dek": "FoundPDE maps observed trajectories to symbolic partial differential equations and exposes where noise breaks recovery.",
    "summary": "FoundPDE maps observed trajectories to symbolic partial differential equations and exposes where noise breaks recovery.",
    "body_text": "FoundPDE turns discretized space-time trajectories into symbolic equations, combining a transformer with Fourier and derivative features plus coefficient refinement. On 20 clean in-domain PDE families, the authors report 97.50 percent type recognition; with a 2 percent multiplicative-noise level, 94.43 percent, falling to 52.09 percent at 10 percent noise. Few-shot tuning recovered more than 90 percent on four held-out composite families. The model is a discovery aid whose equations still require scientific validation.",
    "why_it_matters": "FoundPDE maps observed trajectories to symbolic partial differential equations and exposes where noise breaks recovery.",
    "limitations": [],
    "importance": 9,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-21-008/a-foundation-model-writes-the-equation-behind-the-data",
    "json_url": "https://themachinepress.com/story/mp-2026-10-21-008.json",
    "first_published_at": "2026-10-21T09:00:00.000-04:00",
    "modified_at": "2026-10-21T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-21-008-001",
        "text": "FoundPDE maps observed trajectories to symbolic partial differential equations and exposes where noise breaks recovery.",
        "source_ids": [
          "source-2026-10-21-008"
        ],
        "qualification": null
      }
    ],
    "source_ids": [
      "source-2026-10-21-008"
    ],
    "tags": [
      "PDEs",
      "scientific machine learning",
      "symbolic discovery"
    ],
    "image_url": "https://themachinepress.com/issues/2026-10-21/pde-discovery-lab.webp",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-21-008",
      "title": "arXiv preprint 2610.26963, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.26963v1",
      "canonical_url": "https://arxiv.org/abs/2610.26963v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-20T13:59:00.000-04:00",
      "accessed_at": "2026-10-21T08:27:23.757-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-21-008-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "A Foundation Model Writes the Equation Behind the Data",
    "publisher": "The Machine Press",
    "published_at": "2026-10-21T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-21-008/a-foundation-model-writes-the-equation-behind-the-data"
  }
}