{
  "id": 7169,
  "url": "https://arxiv.org/abs/2603.15397v1",
  "title": "SFCoT: Safer Chain-of-Thought via Active Safety Evaluation and Calibration",
  "summary": "Large language models (LLMs) have demonstrated remarkable capabilities in complex reasoning tasks. However, they remain highly susceptible to jailbreak attacks that undermine their safety alignment. Existing defense mechanisms typically rely on post hoc filtering applied only to the final output, leaving intermediate reasoning steps unmonitored and vulnerable to adversarial manipulation. To address this gap, this paper proposes a SaFer Chain-of-Thought (SFCoT) framework, which proactively evalua",
  "authors": "Yu Pan, Wenlong Yu, Tiejun Wu, Xiaohu Ye, Qiannan Si, Guangquan Xu et al.",
  "category": "research",
  "topics": "safety-alignment,military-security",
  "orgs": null,
  "regions": null,
  "published_at": "2026-03-16T15:13:21.000Z",
  "fetched_at": "2026-07-14T16:33:03.571Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/7169",
  "original_url": "https://arxiv.org/abs/2603.15397v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}