AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Efficient RL Training for LLMs with Experience Replay

Charles Arnal, Vivien Cabannes, Taco Cohen, Julia Kempe, Rémi Munos

ICML 2026 · Reinforcement learning › RL from feedback · approach: RL for LLM reasoning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add an experience replay buffer to on-policy GRPOGRPO · OpenR1-Math-220kcompute to reach target accuracy: save up to 40%Improved
add a replay buffer (accuracy at every compute budget)GRPO · OpenR1-Math-220kaccuracy vs computeImproved
add a replay buffer for training stabilityGRPO · OpenR1-Math-220ktraining stabilityImproved
increase buffer size (ablation)GRPO with replay · Qwen3-0.6B, (W,T)=(6,2)training speed and peak accuracyMixed
lower W/T ratio (raise replay ratio 2.2 to 17.6) (ablation)GRPO with replay · Qwen3-0.6Bmaximal accuracy: 2.2 -> 5.6 -> 17.6Mixed
add replay buffer for pass@kGRPO · Qwen3-0.6Bpass@k for k>1Improved
add moderate off-policiness without a buffer (ablation)GRPO · Qwen3-0.6Btraining stabilityImproved
add positive-bias sampling (delta 0.2 or 0.5) (ablation)GRPO with replay buffer N=4608 · Qwen3-0.6B, OpenR1-Math-220ktest accuracyImproved
replace GRPO with AsymRE loss (ablation)replay buffer N=4608 · Qwen3-0.6B, OpenR1-Math-220ktest accuracyImproved
sample buffer uniformly without replacement (ablation)replay buffer · Qwen3-0.6Btest accuracyNo clear effect
add replay buffer for wall-time speedGRPO async · Qwen3-0.6B / Qwen2.5-7Bwall-time gainImproved
best buffer configs vs best non-buffer configs under learning-rate sweep (ablation)GRPO · OpenR1-Math-220kbest accuracy per computeImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy