| title | Multi-Turn On-Policy Distillation with Prefix Replay |
|---|---|
| authors | Baohao Liao, Hanze Dong, Christof Monz, Xinxing Xu, Li Dong, Furu Wei |
| date | 2026-07-16 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2607.04763 |
| affiliation | Microsoft Research |
| upvotes | 11 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | We study on-policy distillation (OPD) for agentic tasks, where an LLM agent interacts with an environment over multiple turns and a student imitates a teacher over these multi-turn interaction histories. Fully online OPD is costly because each update requires fresh student rollouts through the environment and teacher queries at visited histories. We propose Replayed-Prefix On-Policy Distillation (ReOPD), an off-environment alternative that reuses pre-collected teacher trajectories as replayed prefixes: the student acts at selected steps, while the teacher provides dense per-step supervision without executing new environment interactions. We show that multi-turn OPD introduces a prefix trap: making histories more student-on-policy improves relevance to the student, but can query the teacher on histories where its target is unreliable. This creates a two-sided distribution shift between student occupancy and teacher reliability. ReOPD addresses this by treating multi-turn OPD as a reliability-aware prefix distribution design and implements it with a simple step-decaying sampling schedule that emphasizes early, lower-shift prefixes. Across mathematical reasoning with Python and search environments over multiple teacher and student model scales, ReOPD preserves or improves OPD-level accuracy, uses zero tool calls during student training, and is at least 4times faster per rollout than OPD. ReOPD therefore turns expensive agent-environment interaction into a reusable offline resource, enabling scalable distillation across tools, tasks, and environments. |
cite
@misc{liao2026multi,
title={Multi-Turn On-Policy Distillation with Prefix Replay},
author={Baohao Liao and Hanze Dong and Christof Monz and Xinxing Xu and Li Dong and Furu Wei},
year={2026},
eprint={2607.04763},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2607.04763}
}