{
  "id": 5573,
  "url": "https://arxiv.org/abs/2604.18946v1",
  "title": "Reasoning Structure Matters for Safety Alignment of Reasoning Models",
  "summary": "Large reasoning models (LRMs) achieve strong performance on complex reasoning tasks but often generate harmful responses to malicious user queries. This paper investigates the underlying cause of these safety risks and shows that the issue lies in the reasoning structure itself. Based on this insight, we claim that effective safety alignment can be achieved by altering the reasoning structure. We propose AltTrain, a simple yet effective post training method that explicitly alters the reasoning s",
  "authors": "Yeonjun In, Wonjoong Kim, Sangwu Park, Chanyoung Park",
  "category": "research",
  "topics": "safety-alignment,finance-investment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-21T00:50:13.000Z",
  "fetched_at": "2026-07-14T16:31:53.162Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/5573",
  "original_url": "https://arxiv.org/abs/2604.18946v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}