{
  "id": 14416,
  "url": "https://arxiv.org/abs/2607.26057v1",
  "title": "Pass the Baton: Trajectory-Relayed On-Policy Distillation",
  "summary": "On-policy distillation (OPD) grounds token-level supervision in the student's own trajectory, yet suffers from prefix failure: once the student commits to a wrong reasoning direction, all subsequent generation builds on this deviation, producing misdirected continuations that elicit unreliable supervision and waste compute. We identify a teacher-student continuation asymmetry on failed prefixes, where the teacher tends to redirect while the student continues along the original direction, and con",
  "authors": "Haolei Xu, Xiaowen Xu, Haiwen Hong, Zixuan Ni, Hongxing Li, Yiwen Qiu, Weiming Lu, Yongliang Shen",
  "category": "research",
  "topics": "regulation,children-education",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-28T17:59:46.000Z",
  "fetched_at": "2026-07-29T05:10:12.205Z",
  "source_slug": "x-arxiv-cs-ai",
  "source_name": "arXiv cs.AI",
  "source_homepage": "https://arxiv.org/list/cs.AI/recent",
  "ethics_ai_record_url": "https://ethics.ai/record/14416",
  "original_url": "https://arxiv.org/abs/2607.26057v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}