Pass the Baton: Trajectory-Relayed On-Policy Distillation
The paper introduces Relay On-Policy Distillation (Relay-OPD), a method for on-policy distillation that constructs relay trajectories to address prefix failure and improve performance.
Proposes Relay-OPD to address prefix failure in on-policy distillation
Before reading this…
Applications
- →Mathematical reasoning benchmarks
To understand this paper, make sure you know these concepts first:
- Understanding of on-policy distillationfind papers →
Abstract
More Like ThisOn-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 convert it into a label-free handoff trigger in Relay On-Policy Distillation (Relay-OPD). During training, Relay-OPD constructs relay trajectories by letting the teacher briefly take over at detected trigger points to produce a teacher leg, after which the student resumes and is optimized on the resulting trajectory. A limited relay budget concentrates intervention on critical early positions while limiting departure from the student policy. With a Qwen3-4B-Instruct-2507 teacher and Qwen3-0.6B/1.7B-Non-Thinking students on eight mathematical reasoning benchmarks, Relay-OPD achieves the best or second-best results on every benchmark, outperforming standard OPD by +5.73% and the strongest baseline FastOPD by +1.49% on average for 1.7B, with consistent gains at 0.6B. Training trajectory length is reduced by over 50%.