Preprint
Aug 2026
ReOrder-OPD:Reliability-Aware Prompt Ordering for On-Policy Distillation
On-policy distillation (OPD) applies token-level teacher supervision to student-generated trajectories, but this supervision is not always reliable. Existing methods use local confidence or teacher-student agreement to weight, filter, or truncate the sampled trajectory. These signals do not directly determine whether t...
Ximo Zhu, Rui-Qi Liu, Rong Wang et al.
· 3 citations