When Do We Need On-Policy Distillation? Distilling on Offline Student Rollouts Is Often Better

arXiv:2610.11291 · cs.CL · Submitted 2026-10-08 · Read on arXiv

cs.CL

Submitted: 2026-10-08

Updated: 2026-10-08

Terminology

• Distillation — 133× in this paper · explained in 2 episodes
• On-Policy Distillation — 43× in this paper · explained in 3 episodes
• Accuracy — 27× in this paper · explained in 1 episode
• Alignment — 5× in this paper · explained in 1 episode
• Knowledge Distillation — 4× in this paper · explained in 11 episodes
• AUC — 3× in this paper · explained in 2 episodes
• Exposure Bias — 3× in this paper · explained in 2 episodes
• Failure Modes — 3× in this paper · explained in 3 episodes
• Reinforcement Learning — 3× in this paper · explained in 4 episodes
• Thinking mode — 3× in this paper · explained in 2 episodes
• Benchmark — 2× in this paper · explained in 2 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Core — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Regret — 1× in this paper · explained in 3 episodes
• Supervised Fine-Tuning (SFT) — 1× in this paper · explained in 8 episodes
• Trace — 1× in this paper · explained in 2 episodes

Sources

• DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence
• Demystifying OPD: Length Inflation and Stabilization Strategies for Large Language Models
• When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation
• Revisiting On-Policy Distillation: Empirical Failure Modes and Simple Fixes
• Mismatch Matters: On-Policy Distillation Beyond Token Agreement
• Less is More: Early Stopping Rollout for On-Policy Distillation
• Prefix Teach, Suffix Fade: Local Teachability Collapse in Strong-to-Weak On-Policy Distillation
• Pass the Baton: Trajectory-Relayed On-Policy Distillation
• Trajectory-Refined Distillation
• Kimi K3: Open Frontier Intelligence
• Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation
• Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
• Qwen3 Technical Report
• Nemotron-Cascade 2: Post-Training LLMs with Cascade RL and Multi-Domain On-Policy Distillation
• JustRL: Scaling a 1.5B LLM with a Simple RL Recipe
• A Survey of On-Policy Distillation for Large Language Models
• Entropy-Aware On-Policy Distillation of Language Models
• Unmasking On-Policy Distillation: Where It Helps, Where It Hurts, and Why
• Rethinking On-Policy Distillation of Large Language Models II: One Training Example
• 1% of Tokens Can Be Enough: On Gradient Estimation in On-Policy Distillation

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering