AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Why Tree-Style Branching Matters for Thought Advantage Estimation in GRPO

Hongcheng Wang, Yinuo Huang, Sukai Wang, Guanghui Ren, Hao Dong

ICML 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Adaptive testing and experiments · 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
sample 4 answers per thought (T4A4) instead of 1 (ablation)GRPO T4A1 · GSM8KPass@1: 62.54 -> 70.58Improved
sample 4 answers per thought instead of 1 (ablation)GRPO T4A1 · HumanEvalPass@1: 56.10 -> 58.66Improved
sample 4 answers per thought instead of 1 (ablation)GRPO T4A1 · Trajectory PredictionDFD: 187.99 -> 151.10Improved
answer-level branching (T4A4) instead of scaling thoughts to 16 (T16A1) (ablation)GRPO · Math, Code, vision tasksperformance and cost: GSS@10 15 -> 5 on MathImproved
increase answers per thought from 1 to 4 and 8 (ablation)GRPO-MA · Trajectory PredictionDFD: 187.99 -> 151.10 -> 146.37Improved
remove chain-of-thought (GRPO-NoThink with 16 answers) (ablation)GRPO-MA · Simulated manipulation (seen)success rate: 31.40% -> 10.60%Hurt
use token-length branching (TreeRPO-64) instead of XML-tag branching (ablation)GRPO-MA · Trajectory PredictionDFD: 151.10 -> 157.03Hurt
scale model from 3B to 7B with T8A4 vs T8A1 (ablation)GRPO-MA · Trajectory PredictionDFD: 167.32 -> 134.67Improved
add loss or normalization refinements (DAPO, Dr.GRPO, GRPO-CARE) without branching (ablation)GRPO · Trajectory PredictionDFD: 187.99 -> 180.56 (Dr.GRPO), 188.23 (GRPO-CARE)Mixed

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