AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Long Live The Balance: Information Bottleneck Driven Tree-based Policy Optimization

Hao Jiang, Shurui Li, Tianpeng Bu, Bowen Xu, Xin Liu, Qihua Chen, hongtao duan, lulu hu, Bin Yang, Minying Zhang

ICML 2026 · Reinforcement learning › RL from feedback · approach: GRPO stability and variants · 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 IBTPO (IB-guided tree sampling and IB advantage)GRPO on Qwen3-1.7B-Base · 7 benchmarks overallavg@32 accuracy: 26.3% -> 29.2%Improved
add IBTPOGRPO on Qwen3-8B-Base · 7 benchmarks overallavg@32 accuracy: 40.7% -> 44.3%Improved
add clip-higherGRPO on Qwen3-1.7B-Base · 7 benchmarks overallavg@32 accuracy: 26.3% -> 26.0%No clear effect
add entropy regularizationGRPO on Qwen3-1.7B-Base · 7 benchmarks overallavg@32 accuracy: 26.3% -> 24.9%Hurt
add IBTree sampling (ablation)GRPO on Qwen3-8B-Base · AMC 24avg@32 accuracy: 39.4% -> 43.8%Improved
add IB advantage alone (ablation)GRPO on Qwen3-8B-Base · AIME 25avg@32 accuracy: 13.6% -> 14.2%Improved
add IB advantage to random tree (ablation)RandTree · GPQAavg@32 accuracy: 38.1% -> 37.3%Hurt
replace EPTree with IBTree under IB advantage (ablation)IBTPO · AMC 24avg@32 accuracy: 42.3% -> 46.0%Improved
use IB-Score guided branching instead of entropy guided (ablation)tree sampling G=12 · DAPO-Math-17K subsetAvg-Rate: 21.6% -> 23.2%Improved
use random branching instead of independent sampling (ablation)tree sampling G=12 · DAPO-Math-17K subsetEff-Rate: 59.8% -> 48.4%Hurt
vary beta from 1.0 to 5.0 in IB-Score branching (ablation)IBTree · DAPO-Math-17K subsetEff-Rate: 60.0% -> 60.2%Improved
add local IB advantage with lambda=0.1 to global advantage (ablation)IBTPO · AMC 24avg@32 accuracy: 43.8% -> 46.0%Improved

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