Data-free On-policy Distillation

arXiv:2609.14193 · cs.LG, cs.AI · Submitted 2026-09-12 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-09-12

Updated: 2026-09-17

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Distillation — 50× in this paper · explained in 2 episodes
• On-Policy Distillation — 31× in this paper · explained in 3 episodes
• Benchmark — 10× in this paper · explained in 2 episodes
• Reinforcement Learning — 5× in this paper · explained in 4 episodes
• Accuracy — 3× in this paper · explained in 1 episode
• Alignment — 3× in this paper · explained in 1 episode
• Foundation Model — 3× in this paper · explained in 3 episodes
• Exposure Bias — 2× in this paper · explained in 2 episodes
• Trace — 2× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• GRPO — 1× in this paper · explained in 6 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Generative Model — 1× in this paper · explained in 2 episodes
• Generative Models — 1× in this paper · explained in 2 episodes
• Influence Functions — 1× in this paper · explained in 2 episodes
• Instruction Tuning — 1× in this paper · explained in 2 episodes
• Knowledge Distillation — 1× in this paper · explained in 11 episodes
• Reinforcement Learning (RL) — 1× in this paper · explained in 25 episodes
• Supervised Fine-Tuning (SFT) — 1× in this paper · explained in 8 episodes
• Test-time Scaling — 1× in this paper · explained in 2 episodes
• Thinking mode — 1× in this paper · explained in 2 episodes
• Verifiable Rewards — 1× in this paper · explained in 2 episodes

Sources

• Counteraction-Aware Multi-Teacher On-Policy Distillation for General Capability Recovery with Domain Preservation
• Rethinking On-Policy Distillation of Large Language Models II: One Training Example
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• JustRL: Scaling a 1.5B LLM with a Simple RL Recipe
• Measuring Mathematical Problem Solving With the MATH Dataset
• On-Policy Delta Distillation
• Uni-OPD: Unifying On-Policy Distillation with a Dual-Perspective Recipe
• What Matters in On-Policy Distillation? A Perspective on Data Efficiency and Data Selection
• Editing Models with Task Arithmetic
• R-Diverse: Mitigating Diversity Illusion in Self-Play LLM Training
• On-Policy Distillation with Curriculum Turn-level Guidance for Multi-turn Agents
• LIMR: Less is More for RL Scaling
• UI-MOPD: Multi-Platform On-Policy Distillation for Unified GUI Agents
• MOPD: Multi-Teacher On-Policy Distillation for Capability Integration in LLM Post-Training
• A Survey of On-Policy Distillation for Large Language Models
• SimCT: Recovering Lost Supervision for Cross-Tokenizer On-Policy Distillation
• Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation
• LESS: Selecting Influential Data for Targeted Instruction Tuning
• MiMo-V2-Flash Technical Report

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks