ExGRPO: Learning to Reason from Experience

• URL: http://arxiv.org/abs/2510.02245v1
• Date: Thu, 02 Oct 2025 17:31:30 GMT
• Title: ExGRPO: Learning to Reason from Experience
• Abstract summary: Reinforcement learning from verifiable rewards (RLVR) is an emerging paradigm for improving the reasoning ability of large language models.<n>Standard on-policy training discards rollout experiences after a single update, leading to computational inefficiency and instability.<n>In this paper, we are the first to investigate what makes a reasoning experience valuable and identify rollout correctness and entropy as effective indicators of experience value.
• Score: 82.83309610498446
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement learning from verifiable rewards (RLVR) is an emerging paradigm for improving the reasoning ability of large language models. However, standard on-policy training discards rollout experiences after a single update, leading to computational inefficiency and instability. While prior work on RL has highlighted the benefits of reusing past experience, the role of experience characteristics in shaping learning dynamics of large reasoning models remains underexplored. In this paper, we are the first to investigate what makes a reasoning experience valuable and identify rollout correctness and entropy as effective indicators of experience value. Based on these insights, we propose ExGRPO (Experiential Group Relative Policy Optimization), a framework that organizes and prioritizes valuable experiences, and employs a mixed-policy objective to balance exploration with experience exploitation. Experiments on five backbone models (1.5B-8B parameters) show that ExGRPO consistently improves reasoning performance on mathematical/general benchmarks, with an average gain of +3.5/7.6 points over on-policy RLVR. Moreover, ExGRPO stabilizes training on both stronger and weaker models where on-policy methods fail. These results highlight principled experience management as a key ingredient for efficient and scalable RLVR.

Related papers

• Experience Augmented Policy Optimization for LLM Reasoning [50.16045797892008]
  We argue that experience in RLVR should not be reused as fixed reasoning trajectories, but instead expressed in a policy-adaptive manner.<n>EAPO consistently improves reasoning performance over state-of-the-art RLVR methods.
  arXiv  Detail & Related papers  (2026-06-29T15:05:28Z)
• Learning from Failures: Correction-Oriented Policy Optimization with Verifiable Rewards [73.44333771806282]
  Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective paradigm for improving the reasoning capabilities of large language models.<n>We propose Correction-Oriented Policy Optimization (CIPO), a simple and effective extension to RLVR that converts on-policy failed trajectories into correction-oriented supervision.<n>CIPO improves learning effectiveness while explicitly enhancing the model's ability to correct its own errors.
  arXiv  Detail & Related papers  (2026-05-14T08:22:21Z)
• On Predicting the Post-training Potential of Pre-trained LLMs [60.07459271263409]
  We introduce a new task of predicting post-training potential - forecasting a base model's performance before post-training.<n>We propose RuDE, a unified framework that bypasses the generation gap of base models by leveraging response discrimination.<n>Experiments demonstrate a correlation greater than 90% with post-training performance.
  arXiv  Detail & Related papers  (2026-05-12T11:33:49Z)
• Towards Effective Experiential Learning: Dual Guidance for Utilization and Internalization [71.41478888201401]
  We propose textbfDual textbfGuidance textbfOptimization(textbfDGO) to improve training effectiveness.
  arXiv  Detail & Related papers  (2026-03-25T08:52:56Z)
• Complementary Reinforcement Learning [31.660877399506493]
  Reinforcement Learning (RL) has emerged as a powerful paradigm for training LLM-based agents.<n>Existing approaches suffer from a critical weakness: the experience distilled from history is either stored statically or fail to coevolve with the improving actor.<n>Inspired by complementary learning systems in neuroscience, we present Complementary RL to achieve seamless co-evolution of an experience extractor and a policy actor.
  arXiv  Detail & Related papers  (2026-03-18T11:38:01Z)
• Your Group-Relative Advantage Is Biased [74.57406620907797]
  Group-based learning methods rely on group-relative advantage estimation to avoid learned critics.<n>In this work, we uncover a fundamental issue of group-based RL: the group-relative advantage estimator is inherently biased relative to the true (expected) advantage.<n>We propose History-Aware Adaptive Difficulty Weighting (HA-DW), an adaptive reweighting scheme that adjusts advantage estimates based on an evolving difficulty anchor and training dynamics.
  arXiv  Detail & Related papers  (2026-01-13T13:03:15Z)
• Efficient Reinforcement Learning for Large Language Models with Intrinsic Exploration [33.02780998281276]
  Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning ability of large language models.<n>This study investigates how simply leveraging intrinsic data properties, almost free benefit during training, can improve data efficiency for RLVR.
  arXiv  Detail & Related papers  (2025-11-02T04:16:47Z)
• Beyond Reasoning Gains: Mitigating General Capabilities Forgetting in Large Reasoning Models [33.214586668992965]
  Reinforcement learning with verifiable rewards (RLVR) has delivered impressive gains in mathematical and multimodal reasoning.<n>We propose RECAP-a replay strategy with dynamic objective reweighting for general knowledge.<n>Our method is end-to-end and readily applicable to existing RLVR pipelines without training additional models or heavy tuning.
  arXiv  Detail & Related papers  (2025-10-24T19:08:48Z)
• Attention as a Compass: Efficient Exploration for Process-Supervised RL in Reasoning Models [47.05227816684691]
  We introduce a novel PSRL framework (AttnRL) which enables efficient exploration for reasoning models.<n>Motivated by preliminary observations that steps exhibiting high attention scores correlate with reasoning behaviors, we propose to branch from positions with high values.<n>We develop an adaptive sampling strategy that accounts for problem difficulty and historical batch size, ensuring that the whole training batch maintains non-zero advantage values.
  arXiv  Detail & Related papers  (2025-09-30T17:58:34Z)
• Reinforcement Learning on Pre-Training Data [55.570379963147424]
  We introduce Reinforcement Learning on Pre-Training data (R), a new training-time scaling paradigm for optimizing large language models (LLMs)<n>R enables the policy to autonomously explore meaningful trajectories to learn from pre-training data and improve its capability through reinforcement learning (RL)<n>Extensive experiments on both general-domain and mathematical reasoning benchmarks across multiple models validate the effectiveness of R.
  arXiv  Detail & Related papers  (2025-09-23T17:10:40Z)
• From Trial-and-Error to Improvement: A Systematic Analysis of LLM Exploration Mechanisms in RLVR [92.51110344832178]
  Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for enhancing the reasoning capabilities of large language models (LLMs)<n>This technical report presents a systematic investigation of exploration capacities in RLVR, covering four main aspects.
  arXiv  Detail & Related papers  (2025-08-11T01:26:16Z)
• Perception-Aware Policy Optimization for Multimodal Reasoning [79.56070395437898]
  A major source of error in current multimodal reasoning lies in the perception of visual inputs.<n>We propose PAPO, a novel policy gradient algorithm that encourages the model to learn to perceive while learning to reason.<n>We observe a substantial reduction of 30.5% in perception errors, indicating improved perceptual capabilities with PAPO.
  arXiv  Detail & Related papers  (2025-07-08T23:22:34Z)
• Beyond Accuracy: Dissecting Mathematical Reasoning for LLMs Under Reinforcement Learning [82.43575191712726]
  We introduce a fine-grained analytic framework to dissect the impact ofReinforcement learning on reasoning.<n>Our framework specifically investigates key elements that have been hypothesized to benefit from RL training.
  arXiv  Detail & Related papers  (2025-06-05T07:53:59Z)
• Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback [59.078756231841574]
  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.
  arXiv  Detail & Related papers  (2025-06-03T17:39:02Z)
• Behavior Injection: Preparing Language Models for Reinforcement Learning [24.46625106928253]
  Reinforcement fine-tuning (RFT) has emerged as a powerful post-training technique to incentivize the reasoning ability of large language models (LLMs)<n>LLMs can respond very inconsistently to RFT: some show substantial performance gains, while others plateau or even degrade.<n>We propose behavior injection, a task-agnostic data-augmentation scheme applied prior to RL.
  arXiv  Detail & Related papers  (2025-05-25T00:54:50Z)
• Reward Prediction Error Prioritisation in Experience Replay: The RPE-PER Method [1.600323605807673]
  We introduce Reward Predictive Error Prioritised Experience Replay (RPE-PER)<n>RPE-PER prioritises experiences in the buffer based on RPEs.<n>Our method employs a critic network, EMCN, that predicts rewards in addition to the Q-values produced by standard critic networks.
  arXiv  Detail & Related papers  (2025-01-30T02:09:35Z)
• Conceptual Belief-Informed Reinforcement Learning [10.817700298999]
  Reinforcement learning (RL) has achieved significant success but is hindered by inefficiency and instability.<n>We introduce Conceptual Belief-Informed Reinforcement Learning to emulate human intelligence (HI-RL)<n>HI-RL forms concepts by extracting high-level categories of critical environmental information and then constructs adaptive concept-associated probabilistic beliefs as experience priors to guide value or policy updates.
  arXiv  Detail & Related papers  (2024-10-02T16:50:29Z)
• Iterative Experience Refinement of Software-Developing Agents [81.09737243969758]
  Large language models (LLMs) can leverage past experiences to reduce errors and enhance efficiency. This paper introduces the Iterative Experience Refinement framework, enabling LLM agents to refine experiences iteratively during task execution.
  arXiv  Detail & Related papers  (2024-05-07T11:33:49Z)

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.