AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

Batched Contextual Reinforcement

Bangji Yang, Hongbo Ma, Jiajun Fan, Ge Liu

ICML 2026 · Large language models › Reasoning and chain of thought · approach: RL with verifiable rewards for reasoning

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes Batched Contextual Reinforcement (BCR), a minimalist training approach aimed at reducing chain-of-thought token consumption in LLMs without the quality loss or pipeline complexity of prior efficiency methods.

Problem

Chain-of-Thought reasoning in LLMs consumes excessive tokens, inflating inference cost; existing efficiency methods (explicit length penalties, difficulty estimators, multi-stage curricula) either degrade reasoning quality or require complex training pipelines.

Approach

A minimalist reinforcement approach called Batched Contextual Reinforcement that is described as avoiding explicit length penalties, difficulty estimators, and multi-stage curricula. Details beyond the abstract are not available.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train with batched multi-problem contexts (BCR) vs JustRL, token usageJustRL-deepseek-1.5B · AIME25 (N=1)tokens per problem: 8,482 -> 3,173 (-62.6%)Improved
BCR on JustRL accuracy AIME25JustRL-deepseek-1.5B · AIME25 (N=1)accuracy (%): 40.0 -> 33.3Hurt
BCR on JustRL accuracy MATH-500JustRL-deepseek-1.5B · MATH-500 (N=1)accuracy (%): 91.4 -> 87.6Hurt
BCR on JustRL accuracy AMC23JustRL-deepseek-1.5B · AMC23 (N=1)accuracy (%): 85.0 -> 87.5Improved
BCR on Qwen3-4B-Thinking accuracy AIME25Qwen3-4B-Thinking-2507 · AIME25 (N=1)accuracy (%): 70.0 -> 83.3Improved
BCR on Qwen3-4B-Thinking tokens AIME25Qwen3-4B-Thinking-2507 · AIME25 (N=1)tokens per problem: 20,773 -> 17,498 (-15.8%)Improved
concurrent problems N=1 to N=5 at inference (baseline)JustRL-1.5B · AMC23accuracy (%): 85.0 -> 22.5Hurt
concurrent problems N=1 to N=5 at inference (BCR)BCR · AMC23accuracy (%): 87.5 -> 50.0Hurt
increase concurrent problems N from 1 to 5BCR · AMC23tokens per problem: 2,637 -> 1,013Improved
train with group size N=3 instead of N=4 (ablation)BCR · AMC23 (N=1 eval)accuracy (%): 77.5 -> 87.5Improved
train with group size N=4 instead of N=3 (ablation)BCR · AIME25 (N=1 eval)accuracy (%): 30.0 -> 33.3Improved
train with group size N=3 instead of N=5 (ablation)BCR · Minerva (N=1 eval)accuracy (%): 41.9 -> 46.3Improved

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