arXiv Computation and Language
Bridged Self-Distillation for On-Policy Learning
Read the original on arXiv Computation and Language →The Flow has not summarised this story yet — read it at arXiv Computation and Language.
Related stories
Hugging Face Trending Papers
Oct 19
Bridged Self-Distillation for On-Policy Learning
On-policy distillation can transfer the reasoning capabilities of a strong teacher to a smaller student. However, distillation from a stronger teacher model faces two mismatches: its solutions may be...
arXiv AI
Jun 2
OGLS-SD: On-Policy Self-Distillation with Outcome-Guided Logit Steering for LLM Reasoning
arXiv:2605. 12400v2 Announce Type: replace-cross Abstract: We study on-policy self-distillation (OPSD), where a language model improves its reasoning ability by distilling privileged teacher distributions along its own on-policy trajectories.
arXiv AI
Sep 10
Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why
arXiv:2605.10889v2 Announce Type: replace-cross Abstract: On-policy distillation offers dense, per-token supervision for training reasoning models; however, it remains unclear under which conditions...
arXiv Machine Learning
Jul 30
Weak-to-Strong On-Policy Distillation
arXiv:2607. 26246v1 Announce Type: new Abstract: On-policy distillation (OPD), which aligns a student with the teacher's token-level distribution on the student's own rollouts, is an effective paradigm for transferring capabilities across LLMs.
arXiv AI
Sep 30
Interactive-Policy Distillation with Bidirectional Propose-and-Verify
arXiv:2609.36546v1 Announce Type: cross Abstract: On-policy distillation (OPD) trains a student model on its self-generated trajectories with dense token-level teacher feedback. However, naive OPD ma...
arXiv Machine Learning
Sep 30
Teach Yourself Where to Look: On-Policy Attention Self-Distillation for Reasoning
The paper introduces On-Policy Attention Self-Distillation (OPASD), a method that augments token-level supervision with solution-conditioned attention distillation for reasoning models. OPASD projects a privileged teacher’s attention onto student-visible positions, renormalizes the distribution, and aligns it with the student. Experiments on three model sizes and four math benchmarks show that OPASD improves accuracy by 4.98–8.40 percentage points, reduces generated tokens by 73.9%, cuts compute by 72.6%, and trains 1.53× faster compared to token-only distillation.