REVO: Rollout-Efficient Off-Policy Distillation via Variance-Guided Reuse

arXiv:2609.37500 · cs.LG · Submitted 2026-09-28 · Read on arXiv

cs.LG

Submitted: 2026-09-28

Updated: 2026-09-28

Code: https://github.com/UNCSciML/REVO

Terminology

• Distillation — 1× in this paper · explained in 2 episodes

Sources

• Efficient RL Training for LLMs with Experience Replay
• OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problems
• Distilling the Knowledge in a Neural Network
• Entropy-Aware On-Policy Distillation of Language Models
• AsyncOPD: How Stale Can On-Policy Distillation Be?
• Filter, Then Reweight: Rethinking Optimization Granularity in On-Policy Distillation
• GPQA: A Graduate-Level Google-Proof Q&A Benchmark
• Kimi K3: Open Frontier Intelligence
• Qwen3 Technical Report
• Self-Distilled RLVR
• OGLS-SD: On-Policy Self-Distillation with Outcome-Guided Logit Steering for LLM Reasoning
• When EOS Tokens Disagree: Understanding Length Inflation in On-Policy Distillation
• DAPO: An Open-Source LLM Reinforcement Learning System at Scale
• Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective
• Self-Distilled Reasoner: On-Policy Self-Distillation for Large Language Models
• Group Sequence Policy Optimization
• Less is More: Early Stopping Rollout for On-Policy Distillation

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: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks