Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback

• URL: http://arxiv.org/abs/2506.03106v4
• Date: Thu, 17 Jul 2025 04:08:03 GMT
• Title: Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback
• Authors: Xiaoying Zhang, Hao Sun, Yipeng Zhang, Kaituo Feng, Chaochao Lu, Chao Yang, Helen Meng,
• Abstract summary: Critique-GRPO is an online RL framework that integrates both natural language and numerical feedback for effective policy optimization.<n>We show Critique-GRPO consistently outperforms supervised learning and RL-based fine-tuning methods across eight challenging mathematical, STEM, and general reasoning tasks.
• Score: 59.078756231841574
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in reinforcement learning (RL) with numerical feedback, such as scalar rewards, have significantly enhanced the complex reasoning capabilities of large language models (LLMs). Despite this success, we identify three key challenges encountered by RL with solely numerical feedback: performance plateaus, limited effectiveness of spontaneous self-reflection, and persistent failures. We then demonstrate that RL-finetuned models, even after exhibiting performance plateaus, can generate correct refinements on persistently failed problems by leveraging natural language feedback in the form of critiques. Building on this insight, we propose Critique-GRPO, an online RL framework that integrates both natural language and numerical feedback for effective policy optimization. Critique-GRPO enables LLMs to learn from initial responses and critique-guided self-refinements simultaneously while maintaining exploration. Additionally, we employ a shaping function to amplify learning from correct, especially unfamiliar, refinements and penalize incorrect ones. Extensive experiments with Qwen2.5-7B-Base, Qwen2.5-Math-7B-Base, and Qwen3-8B demonstrate that Critique-GRPO consistently outperforms supervised learning and RL-based fine-tuning methods across eight challenging mathematical, STEM, and general reasoning tasks, improving average pass@1 scores by approximately 4.4% and 3.8% on Qwen2.5-7B-Base and Qwen3-8B, respectively. Notably, Critique-GRPO enables effective self-improvement through self-critiquing and weak-to-strong generalization, achieving consistent gains over GRPO, such as 16.7% and 10.0% pass@1 improvements on AIME 2024, respectively.

Related papers

• Trustworthy Reasoning: Evaluating and Enhancing Factual Accuracy in LLM Intermediate Thought Processes [16.451488374845407]
  We present a novel framework addressing a critical vulnerability in Large Language Models (LLMs)<n>This phenomenon poses substantial risks in high-stakes domains including healthcare, legal analysis, and scientific research.
  arXiv  Detail & Related papers  (2025-07-25T10:34:51Z)
• RefCritic: Training Long Chain-of-Thought Critic Models with Refinement Feedback [57.967762383794806]
  RefCritic is a long-chain-of-thought critic module based on reinforcement learning with dual rule-based rewards.<n>We evaluate RefCritic on Qwen2.5-14B-Instruct and DeepSeek-R1-Distill-Qwen-14B across five benchmarks.
  arXiv  Detail & Related papers  (2025-07-20T16:19:51Z)
• GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark with complex real-world videos requiring balanced perception and reasoning.<n>Using SEED-Bench-R1, we find that standard GRPO, while improving answer accuracy, often reduces logical coherence between reasoning steps and answers, with only a 57.9% consistency rate.<n>We propose GRPO-CARE, a consistency-aware RL framework optimizing both answer correctness and reasoning coherence without explicit supervision.
  arXiv  Detail & Related papers  (2025-06-19T08:49:13Z)
• Boosting LLM Reasoning via Spontaneous Self-Correction [43.4980625253775]
  One of the approaches for improving math reasoning is self-correction.<n>Existing self-correction approaches treat corrections as standalone post-generation refinements.<n>We propose SPOC, a spontaneous self-correction approach that enables LLMs to generate interleaved solutions and verifications in a single inference pass.
  arXiv  Detail & Related papers  (2025-06-07T21:23:00Z)
• Navigate the Unknown: Enhancing LLM Reasoning with Intrinsic Motivation Guided Exploration [33.807927649100805]
  Reinforcement learning (RL) has emerged as a pivotal method for improving the reasoning capabilities of Large Language Models (LLMs)<n>RL approaches face critical limitations due to their reliance on sparse outcome-based rewards and inadequate mechanisms for incentivizing exploration.<n>We propose an Intrinsic Motivation guidEd exploratioN meThOd foR LLM Reasoning (i-MENTOR)<n>i-MENTOR introduces three key innovations: trajectory-aware exploration rewards that mitigate bias in token-level strategies; dynamic reward scaling to stabilize exploration and exploitation in large action spaces; and advantage-preserving reward implementation that maintains
  arXiv  Detail & Related papers  (2025-05-23T08:30:28Z)
• SRPO: A Cross-Domain Implementation of Large-Scale Reinforcement Learning on LLM [18.275547804539016]
  Two-Staged history-Resampling Policy optimization surpasses the performance of DeepSeek-R1-Zero-32B on the AIME24 and LiveCodeBench benchmarks.<n>We introduce two key methodological innovations: (1) a two-stage cross-domain training paradigm designed to balance the development of mathematical reasoning and coding proficiency, and (2) History Resampling (HR), a technique to address ineffective samples.
  arXiv  Detail & Related papers  (2025-04-19T13:06:03Z)
• Nemotron-CrossThink: Scaling Self-Learning beyond Math Reasoning [66.43194385702297]
  Large Language Models (LLMs) have shown strong reasoning capabilities, particularly when enhanced through Reinforcement Learning (RL)<n>We propose NEMOTRON-CROSSTHINK, a framework that systematically incorporates multi-domain corpora, including both synthetic and real-world question-answer pairs, into RL training to improve generalization across diverse reasoning tasks.
  arXiv  Detail & Related papers  (2025-04-15T21:37:13Z)
• GRPO-LEAD: A Difficulty-Aware Reinforcement Learning Approach for Concise Mathematical Reasoning in Language Models [0.17265013728931003]
  GRPO-LEAD is a suite of novel enhancements tailored for mathematical reasoning.<n>It introduces (1) a length-dependent accuracy reward to encourage concise and precise solutions, (2) an explicit penalty mechanism for incorrect answers to sharpen decision boundaries, and (3) a difficulty-aware advantage reweighting strategy that amplifies learning signals for challenging problems.
  arXiv  Detail & Related papers  (2025-04-13T19:07:45Z)
• R-PRM: Reasoning-Driven Process Reward Modeling [53.06844294668382]
  Process Reward Models (PRMs) have emerged as a promising solution by evaluating each reasoning step.<n>Existing PRMs typically output evaluation scores directly, limiting both learning efficiency and evaluation accuracy.<n>We propose Reasoning-Driven Process Reward Modeling (R-PRM)<n>R-PRM generates seed data from limited annotations, effectively bootstrapping our model's reasoning capabilities.
  arXiv  Detail & Related papers  (2025-03-27T09:23:08Z)
• OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement [91.88062410741833]
  This study investigates whether similar reasoning capabilities can be successfully integrated into large vision-language models (LVLMs)<n>We consider an approach that iteratively leverages supervised fine-tuning (SFT) on lightweight training data and Reinforcement Learning (RL) to further improve model generalization.<n>OpenVLThinker, a LVLM exhibiting consistently improved reasoning performance on challenging benchmarks such as MathVista, MathVerse, and MathVision, demonstrates the potential of our strategy for robust vision-language reasoning.
  arXiv  Detail & Related papers  (2025-03-21T17:52:43Z)
• Self-Evolving Critique Abilities in Large Language Models [59.861013614500024]
  This paper explores enhancing critique abilities of Large Language Models (LLMs)<n>We introduce SCRIT, a framework that trains LLMs with self-generated data to evolve their critique abilities.<n>Our analysis reveals that SCRIT's performance scales positively with data and model size.
  arXiv  Detail & Related papers  (2025-01-10T05:51:52Z)
• Improving Multi-Step Reasoning Abilities of Large Language Models with Direct Advantage Policy Optimization [22.67700436936984]
  We introduce Direct Advantage Policy Optimization (DAPO), a novel step-level offline reinforcement learning algorithm.<n>DAPO employs a critic function to predict the reasoning accuracy at each step, thereby generating dense signals to refine the generation strategy.<n>Our results show that DAPO can effectively enhance the mathematical and code capabilities on both SFT models and RL models, demonstrating the effectiveness of DAPO.
  arXiv  Detail & Related papers  (2024-12-24T08:39:35Z)
• Self-Generated Critiques Boost Reward Modeling for Language Models [57.60881438647227]
  Critic-RM is a framework that improves reward models using self-generated critiques without extra supervision.<n> Experiments show that Critic-RM improves reward modeling accuracy by 3.7%-7.3% compared to standard reward models and LLM judges.
  arXiv  Detail & Related papers  (2024-11-25T18:28:26Z)
• Training Language Models to Critique With Multi-agent Feedback [102.42751835338233]
  MultiCritique pipeline improves critique ability of LLMs by utilizing multi-agent feedback. pipeline aggregates high-quality critiques from multiple agents instead of a single model. Our fine-tuned 7B model significantly surpasses other advanced 7B-13B open-source models.
  arXiv  Detail & Related papers  (2024-10-20T04:57:45Z)
• A Comprehensive Evaluation of Large Language Models on Mental Illnesses [0.8458496687170665]
  GPT-4 and Llama 3 exhibited superior performance in binary disorder detection, with accuracies reaching up to 85% on certain datasets. prompt engineering played a crucial role in enhancing model performance. Despite promising results, our analysis identified several challenges, including variability in performance across datasets and the need for careful prompt engineering.
  arXiv  Detail & Related papers  (2024-09-24T02:58:52Z)
• Learning to Refine with Fine-Grained Natural Language Feedback [81.70313509881315]
  We propose looking at refinement with feedback as a composition of three distinct LLM competencies. A key property of the proposed Detect, Critique, Refine ("DCR") method is that the step 2 critique model can give fine-grained feedback about errors. We show that models of different capabilities benefit from refining with DCR on the task of improving factual consistency of document grounded summaries.
  arXiv  Detail & Related papers  (2024-07-02T16:15:01Z)
• Towards Reliable and Fluent Large Language Models: Incorporating Feedback Learning Loops in QA Systems [10.58737969057445]
  We build a dataset to train a critic model capable of evaluating the citation, correctness, and fluency of responses generated by large language models. We propose an automated feedback mechanism that leverages the critic model to offer real-time feedback on heterogeneous aspects of generated text. Experimental results demonstrate the efficacy of our approach, including a 4% precision increase in citation and an approximately 8% enhancement in the MAUVE metric for fluency.
  arXiv  Detail & Related papers  (2023-09-08T09:39:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.