{
  "url": "https://appliedfrontier.nl/en/research/signal-210b5a5aac34c71d/",
  "title": "INSPIRE: An Internalize-Then-Improve Approach for Example-Driven Mathematical Reasoning",
  "published": "2026-08-27",
  "excerpt": "Mathematical reasoning has seen rapid progress in large language models (LLMs), yet existing methods optimize predominantly for final-answer correctness, raising the question whether models truly",
  "content_sha256": "a5f572da96074feed017c6ee13d1df2d58e1da9c650237db4ec8f61728770928",
  "extracted_chars": 1425,
  "artifact_links": [
    "https://arxiv.org/abs/2608.27501"
  ],
  "arxiv_id": "2608.27501",
  "id": "signal-210b5a5aac34c71d",
  "org": "Research community",
  "topic": "Mathematical reasoning",
  "source_kind": "preprint",
  "status": "unreviewed-source-report",
  "analysis_method": "deterministic_metadata_triage_v1",
  "novelty_verified": false,
  "proof_checked": false,
  "peer_review_verified": false,
  "matched_terms": [
    "mathematical",
    "mathematics",
    "reasoning"
  ],
  "checked_at": "2026-10-29T22:08:08+00:00",
  "first_seen": "2026-10-29T21:52:00+00:00",
  "source_hash": "b7343bad585a6109cf365fc8f4b1649af65e5aaecc385238b3c7e2535b32c709",
  "source_url": "https://arxiv.org/abs/2608.27501",
  "language": "en",
  "indexing": "noindex_until_editorially_enriched"
}