Shu Zhao, Tan Yu · 2026 · arXiv
DOI: 10.48550/arXiv.2610.28586 · open-access PDF
Agentic on-policy distillation (OPD) melatih agen siswa pada jalurnya sendiri sementara seorang guru mencetak setiap token. Penjelasan alami tentang di mana hal itu salah adalah kesalahan komposisi: kesalahan siswa membawa jalurnya melampaui wilayah di mana guru dapat diandalkan, sehingga metode terbaru terus melatih di wilayah itu, melalui kurikulum panjang episode, putaran yang dipandu oleh guru, replay prefix, atau intervensi guru di mana keduanya tidak setuju. Kami menunjukkan penjelasan ini tidak lengkap.
Machine translation (Meta M2M100); the English original follows.
Agentic on-policy distillation (OPD) trains a student agent on its own trajectories while a teacher scores every token. A natural explanation of where it goes wrong is compounding error: the student's mistakes carry its trajectories beyond the region where the teacher is reliable, so recent methods keep training inside that region, through episode-length curricula, teacher-guided turns, prefix replay, or teacher intervention where the two disagree. We show this explanation is incomplete. From the states where the student departs most from the teacher, the teacher's own takeovers still finish t
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28586, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange