{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-08-27-014",
    "source_story_id": "tmp-story-finriskatlas",
    "edition_id": "mp-2026-08-27-morning-0049",
    "edition_url": "https://themachinepress.com/edition/2026-08-27",
    "position": 14,
    "story_type": "dispatch",
    "section": "benchmarks-evals",
    "editorial_classification": "editorial",
    "headline": "The Broad Finance Score Hid Decision-Level Regret",
    "slug": "the-broad-finance-score-hid-decision-level-regret",
    "dek": "FinRiskAtlas evaluates specific review operations and whether the available evidence supports a defensible next step.",
    "summary": "FinRiskAtlas evaluates specific review operations and whether the available evidence supports a defensible next step.",
    "body_text": "The Chinese-language suite contains 9,742 static instances across 53 task families plus 680 replayed pre-action states from 104 de-identified trajectories. Across 33 model configurations, operation rankings correlated only 0.42 on average, and knowledge-based shortlisting incurred up to 18.01 points of regret on individual operations.",
    "why_it_matters": "FinRiskAtlas evaluates specific review operations and whether the available evidence supports a defensible next step.",
    "limitations": [
      "Across 33 model configurations, operation rankings correlated only 0.42 on average, and knowledge-based shortlisting incurred up to 18.01 points of regret on individual operations."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-27-014/the-broad-finance-score-hid-decision-level-regret",
    "json_url": "https://themachinepress.com/story/mp-2026-08-27-014.json",
    "first_published_at": "2026-08-27T09:00:00.000-04:00",
    "modified_at": "2026-08-27T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-08-27-014-001",
        "text": "FinRiskAtlas evaluates specific review operations and whether the available evidence supports a defensible next step.",
        "source_ids": [
          "source-2026-08-27-014"
        ],
        "qualification": "Across 33 model configurations, operation rankings correlated only 0.42 on average, and knowledge-based shortlisting incurred up to 18.01 points of regret on individual operations."
      }
    ],
    "source_ids": [
      "source-2026-08-27-014"
    ],
    "tags": [
      "financial AI",
      "evidence states",
      "professional review"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-08-27-014",
      "title": "arXiv preprint 2608.25325",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2608.25325",
      "canonical_url": "https://arxiv.org/abs/2608.25325",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-08-25T23:13:15.000-04:00",
      "accessed_at": "2026-08-27T08:18:00.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-08-27-014-001"
      ]
    }
  ],
  "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 Broad Finance Score Hid Decision-Level Regret",
    "publisher": "The Machine Press",
    "published_at": "2026-08-27T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-27-014/the-broad-finance-score-hid-decision-level-regret"
  }
}