Cover Image

Step-GRPO: Internalizing Dynamic Early Exit for Efficient Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Chen, Benteng, Wang, Weida, Zhang, Shufei, Lin, Mingbao, Zhang, Min
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

S-GRPO: Early Exit via Reinforcement Learning in Reasoning Models
by: Dai, Muzhi, et al.
Published: (2025)

DSADF: Thinking Fast and Slow for Decision Making
by: Dou, Zhihao, et al.
Published: (2025)

Dynamic Early Exit in Reasoning Models
by: Yang, Chenxu, et al.
Published: (2025)

Plan Then Action:High-Level Planning Guidance Reinforcement Learning for LLM Reasoning
by: Dou, Zhihao, et al.
Published: (2025)

FlashThink: An Early Exit Method For Efficient Reasoning
by: Jiang, Guochao, et al.
Published: (2025)

The Zero-Step Thinking: An Empirical Study of Mode Selection as Harder Early Exit in Reasoning Models
by: Tan, Yuqiao, et al.
Published: (2025)

CPPO: Accelerating the Training of Group Relative Policy Optimization-Based Reasoning Models
by: Lin, Zhihang, et al.
Published: (2025)

SCOPE: Compress Mathematical Reasoning Steps for Efficient Automated Process Annotation
by: Xu, Huimin, et al.
Published: (2025)

ADEPT: Adaptive Dynamic Early-Exit Process for Transformers
by: Yoo, Sangmin, et al.
Published: (2026)

Dynamic Vocabulary Pruning in Early-Exit LLMs
by: Vincenti, Jort, et al.
Published: (2024)

SpecExit: Accelerating Large Reasoning Model via Speculative Exit
by: Yang, Rubing, et al.
Published: (2025)

DaGRPO: Rectifying Gradient Conflict in Reasoning via Distinctiveness-Aware Group Relative Policy Optimization
by: Xie, Xuan, et al.
Published: (2025)

E4: Energy-Efficient DNN Inference for Edge Video Analytics Via Early-Exit and DVFS
by: Zhang, Ziyang, et al.
Published: (2025)

Early-Exit Neural Networks with Nested Prediction Sets
by: Jazbec, Metod, et al.
Published: (2023)

SEED-GRPO: Semantic Entropy Enhanced GRPO for Uncertainty-Aware Policy Optimization
by: Chen, Minghan, et al.
Published: (2025)

GRPO and Reflection Reward for Mathematical Reasoning in Large Language Models
by: Wang, Zhijie
Published: (2026)

BranchGRPO: Stable and Efficient GRPO with Structured Branching in Diffusion Models
by: Li, Yuming, et al.
Published: (2025)

Internalizing Safety Understanding in Large Reasoning Models via Verification
by: Zhang, Yi, et al.
Published: (2026)

Beyond Greedy Exits: Improved Early Exit Decisions for Risk Control and Reliability
by: Bajpai, Divya Jyoti, et al.
Published: (2025)

Unbiased Dynamic Pruning for Efficient Group-Based Policy Optimization
by: Zhu, Haodong, et al.
Published: (2026)

LYNX: Learning Dynamic Exits for Confidence-Controlled Reasoning
by: Akgül, Ömer Faruk, et al.
Published: (2025)

EDGE-GRPO: Entropy-Driven GRPO with Guided Error Correction for Advantage Diversity
by: Zhang, Xingjian, et al.
Published: (2025)

Reversible Diffusion Decoding for Diffusion Language Models
by: Wang, Xinyun, et al.
Published: (2026)

Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback
by: Zhang, Xiaoying, et al.
Published: (2025)

QCBench: Evaluating Large Language Models on Domain-Specific Quantitative Chemistry
by: Xie, Jiaqing, et al.
Published: (2025)

Stepwise Guided Policy Optimization: Coloring your Incorrect Reasoning in GRPO
by: Chen, Peter, et al.
Published: (2025)

E-GRPO: High Entropy Steps Drive Effective Reinforcement Learning for Flow Models
by: Zhang, Shengjun, et al.
Published: (2026)

ReProbe: Efficient Test-Time Scaling of Multi-Step Reasoning by Probing Internal States of Large Language Models
by: Ni, Jingwei, et al.
Published: (2025)

The Diminishing Returns of Early-Exit Decoding in Modern LLMs
by: Wei, Rui, et al.
Published: (2026)

iGRPO: Self-Feedback-Driven LLM Reasoning
by: Hatamizadeh, Ali, et al.
Published: (2026)

GRPO-$λ$: Credit Assignment improves LLM Reasoning
by: Parthasarathi, Prasanna, et al.
Published: (2025)

Amortized-Precision Quantization for Early-Exit Vision Transformers
by: Fang, Rui, et al.
Published: (2026)

TERMINATOR: Learning Optimal Exit Points for Early Stopping in Chain-of-Thought Reasoning
by: Nagle, Alliot, et al.
Published: (2026)

DeeAD: Dynamic Early Exit of Vision-Language Action for Efficient Autonomous Driving
by: HU, Haibo, et al.
Published: (2025)

Normalized Narrow Jump To Conclusions: Normalized Narrow Shortcuts for Parameter Efficient Early Exit Transformer Prediction
by: Seshadri, Amrit Diggavi
Published: (2024)

DIVA-GRPO: Enhancing Multimodal Reasoning through Difficulty-Adaptive Variant Advantage
by: Gao, Haowen, et al.
Published: (2026)

Temporal Decisions: Leveraging Temporal Correlation for Efficient Decisions in Early Exit Neural Networks
by: Sponner, Max, et al.
Published: (2024)

Seeing World Dynamics in a Nutshell
by: Shen, Qiuhong, et al.
Published: (2025)

ESTAR: Early-Stopping Token-Aware Reasoning For Efficient Inference
by: Wang, Junda, et al.
Published: (2026)

Advancing Multimodal Large Language Models with Quantization-Aware Scale Learning for Efficient Adaptation
by: Xie, Jingjing, et al.
Published: (2024)