Reinforcement Learning with Verifiable Rewards (RLVR) and GRPO for Reasoning

A technical note on Reinforcement Learning with Verifiable Rewards (RLVR) and Group Relative Policy Optimization (GRPO), and how these methods are used to improve multi-step reasoning capabilities in language models.

Reinforcement Learning in Robotics Speech and dialogue systems Natural Language Processing Techniques

Source

Paper

The full text of this publication is not hosted on 44B due to licensing.

Read it at OpenAlex

GRPO-VPS: Enhancing Group Relative Policy Optimization with Verifiable Process Supervision for Effective ReasoningJingyi Wang, Lei Zhu, Tengjin Weng et al.2026 · 1 citation In Library

Explore Data Left Behind in Reinforcement Learning for Reasoning Language ModelsChenxi Liu, Junjie Liang, Yuqi Jia et al.2025 · arXiv.org · 6 citations In Library

Revisiting Reinforcement Learning with Verifiable Rewards from a Contrastive PerspectiveFeng Zhang, Xinhong Ma, Ziqiang Dong et al.2026 · arXiv.org · 0 citations In Library

Off-Context GRPO: Learning to Reason on Hard Problems using Privileged InformationPriyank Agrawal, Ankur Samanta, S. Ghasemlou et al.2026 · 0 citations In Library

Can GRPO Help LLMs Transcend Their Pretraining Origin?Kangqi Ni, Zhen Tan, Zijie Liu et al.2025 · arXiv.org · 5 citations In Library

Structured Role-Aware Policy Optimization for Multimodal ReasoningBin Jiang, Difan Zou2026 · 0 citations In Library

iGRPO: Self-Feedback-Driven LLM ReasoningAli Hatamizadeh, Shrimai Prabhumoye, Igor Gitman et al.2026 · arXiv.org · 1 citation In Library

R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy OptimizationJingyi Zhang, Jiaxing Huang, Huanjin Yao et al.2025 · IEEE International Conference on Computer Vision · 326 citations In Library

Risk-Sensitive RL for Alleviating Exploration Dilemmas in Large Language ModelsYuhua Jiang, Jiawei Huang, Yufeng Yuan et al.2025 · arXiv.org · 11 citations In Library

Patents near this paper

SYSTEMS AND METHODS FOR LEARNING REUSABLE OPTIONS TO TRANSFER KNOWLEDGE BETWEEN TASKSUS 11,511,413 · HUAWEI TECHNOLOIES CO., LTD. · 2022 SYSTEMS AND METHODS FOR END-TO-END DEEP REINFORCEMENT LEARNING BASED COREFERENCE RESOLUTIONUS 11,630,953 · BAIDU USA LLC · 2023 SECURE EXPLORATION FOR REINFORCEMENT LEARNINGUS 11,616,813 · MICROSOFT TECHNOLOGY LICENSING, LLC · 2023 LEXICOGRAPHIC DEEP REINFORCEMENT LEARNING USING STATE CONSTRAINTS AND CONDITIONAL POLICIESUS 11,410,023 · INTERNATIONAL BUSINESS MACHINES CORPORATION · 2022 Patent US 11,836,590US 11,836,590

Source of record

OpenAlex

Reinforcement Learning with Verifiable Rewards (RLVR) and GRPO for Reasoning

Paper

Similar papers

Patents near this paper

Welcome.

Patents near this paper