MMLU (easier) shows significantly more performative reasoning than GPQA-Diamond (harder).

{
  "doc": "boppana-goodfire-reasoning-theater-2026.md",
  "author": null,
  "context": "Core interpretive claim: task difficulty correlates inversely with amount of performative (post-hoc) reasoning.",
  "enrichment": {
    "is_stale": true
  },
  "norm_label": "MMLU (easier) shows significantly more performative reasoning than GPQA-Diamond (harder).",
  "source_url": null,
  "graphify_id": "claim_easier_tasks_more_performative",
  "source_file": "boppana-goodfire-reasoning-theater-2026.md",
  "imported_from": "/Users/antonborzov/Documents/Research.nosync/papers/extract_typed_out/boppana-goodfire-reasoning-theater-2026/graph.json",
  "extracted_type": "claim",
  "graphify_file_type": "claim"
}