R^3: Replay, Reflection, and Ranking Rewards for LLM Reinforcement Learning

• URL: http://arxiv.org/abs/2601.19620v2
• Date: Wed, 28 Jan 2026 02:38:20 GMT
• Title: R^3: Replay, Reflection, and Ranking Rewards for LLM Reinforcement Learning
• Authors: Zhizheng Jiang, Kang Zhao, Weikai Xu, Xinkui Lin, Wei Liu, Jian Luan, Shuo Shang, Peng Han,
• Abstract summary: Large reasoning models (LRMs) aim to solve diverse and complex problems through structured reasoning.<n>Recent advances in group-based policy optimization methods have shown promise in enabling stable advantage estimation without reliance on process-level annotations.<n>We propose a reinforcement learning mechanism named emphtextbfR3 that along three directions: (1) a emphcross-context underlinetextbfReplay strategy that maintains the intra-group advantage, (2) an emphin-context self-underlinetextbfReflection mechanism
• Score: 32.16683059021539
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large reasoning models (LRMs) aim to solve diverse and complex problems through structured reasoning. Recent advances in group-based policy optimization methods have shown promise in enabling stable advantage estimation without reliance on process-level annotations. However, these methods rely on advantage gaps induced by high-quality samples within the same batch, which makes the training process fragile and inefficient when intra-group advantages collapse under challenging tasks. To address these problems, we propose a reinforcement learning mechanism named \emph{\textbf{R^3}} that along three directions: (1) a \emph{cross-context \underline{\textbf{R}}eplay} strategy that maintains the intra-group advantage by recalling valuable examples from historical trajectories of the same query, (2) an \emph{in-context self-\underline{\textbf{R}}eflection} mechanism enabling models to refine outputs by leveraging past failures, and (3) a \emph{structural entropy \underline{\textbf{R}}anking reward}, which assigns relative rewards to truncated or failed samples by ranking responses based on token-level entropy patterns, capturing both local exploration and global stability. We implement our method on Deepseek-R1-Distill-Qwen-1.5B and train it on the DeepscaleR-40k in the math domain. Experiments demonstrate our method achieves SoTA performance on several math benchmarks, representing significant improvements and fewer reasoning tokens over the base models. Code and model will be released.

Related papers

• TopoCurate:Modeling Interaction Topology for Tool-Use Agent Training [53.93696896939915]
  Training tool-use agents typically rely on Supervised Fine-Tuning (SFT) on successful trajectories and Reinforcement Learning (RL) on pass-rate-selected tasks.<n>We propose TopoCurate, an interaction-aware framework that projects multi-trial rollouts from the same task into a unified semantic quotient topology.<n>TopoCurate achieves consistent gains of 4.2% (SFT) and 6.9% (RL) over state-of-the-art baselines.
  arXiv  Detail & Related papers  (2026-03-02T10:38:54Z)
• Towards Sample-Efficient and Stable Reinforcement Learning for LLM-based Recommendation [56.92367609590823]
  Long Chain-of-Thought (Long CoT) reasoning has shown promise in Large Language Models (LLMs)<n>We argue that Long CoT is inherently ill-suited for the sequential recommendation domain.<n>We propose RISER, a novel Reinforced Item Space Exploration framework for Recommendation.
  arXiv  Detail & Related papers  (2026-01-31T10:02:43Z)
• VADE: Variance-Aware Dynamic Sampling via Online Sample-Level Difficulty Estimation for Multimodal RL [38.782188833641676]
  Group-based policy optimization methods like GRPO and GSPO have become standard for training multimodal models.<n>They suffer from a critical emphgradient vanishing problem when all responses within a group receive identical rewards.<n>We propose textbfVADE, a sampling framework via online sample-level difficulty textbfEstimation.
  arXiv  Detail & Related papers  (2025-11-24T08:59:54Z)
• Scaf-GRPO: Scaffolded Group Relative Policy Optimization for Enhancing LLM Reasoning [49.290631188365786]
  Scaf-GRPO is a training framework that intervenes when a model's independent learning has plateaued.<n>It boosts the pass@1 score of the Qwen2.5-Math-7B model by a relative 44.3% over a vanilla GRPO baseline.<n>This result demonstrates our framework provides a robust and effective methodology for unlocking a model's ability to solve problems previously beyond its reach.
  arXiv  Detail & Related papers  (2025-10-22T17:41:30Z)
• LLM Routing with Dueling Feedback [49.67815163970033]
  We study the problem of selecting the best model for each query while balancing user satisfaction, model expertise, and inference cost.<n>We formulate routing as contextual dueling bandits, learning from pairwise preference feedback rather than absolute scores.<n>We introduce Category-Calibrated Fine-Tuning (CCFT), a representation-learning method that derives model embeddings from offline data using contrastive fine-tuning with categorical weighting.
  arXiv  Detail & Related papers  (2025-10-01T12:52:25Z)
• Nested-ReFT: Efficient Reinforcement Learning for Large Language Model Fine-Tuning via Off-Policy Rollouts [25.205293698698867]
  We introduce Nested-ReFT, where a subset of layers of the target model acts as the behavior model to generate off-policy completions during training.<n>Our theoretical analysis shows that Nested-ReFT yields unbiased gradient estimates with controlled variance.<n>Our empirical analysis demonstrates improved computational efficiency measured as tokens/sec across multiple math reasoning benchmarks and model sizes.
  arXiv  Detail & Related papers  (2025-08-13T18:37:46Z)
• From Sufficiency to Reflection: Reinforcement-Guided Thinking Quality in Retrieval-Augmented Reasoning for LLMs [13.410543801811992]
  This paper analyzes existing RAG reasoning models and identifies three main failure patterns.<n>We propose TIRESRAG-R1, a novel framework using a think-retrieve-reflect process and a multi-dimensional reward system.<n>Experiments on four multi-hop QA datasets show that TIRESRAG-R1 outperforms prior RAG methods and generalizes well to single-hop tasks.
  arXiv  Detail & Related papers  (2025-07-30T14:29:44Z)
• Reinforced Latent Reasoning for LLM-based Recommendation [92.56166822197919]
  Large Language Models (LLMs) have demonstrated impressive reasoning capabilities in complex problem-solving tasks.<n>Existing methods typically rely on fine-tuning with explicit chain-of-thought (CoT) data.<n>In this work, we explore an alternative approach that shifts from explicit CoT reasoning to compact, information-dense latent reasoning.
  arXiv  Detail & Related papers  (2025-05-25T11:03:45Z)
• Unearthing Gems from Stones: Policy Optimization with Negative Sample Augmentation for LLM Reasoning [41.83677588934301]
  We propose Behavior Constrained Policy Gradient with Negative Sample Augmentation (BCPG-NSA)<n>BCPG-NSA is a fine-grained offline framework that encompasses three stages: 1) sample segmentation, 2) consensus-based step correctness assessment combining LLM and PRM judgers, and 3) policy optimization with NSA designed to effectively mine positive steps within negative samples.<n> Experimental results show that BCPG-NSA outperforms baselines on several challenging math/coding reasoning benchmarks using the same training dataset.
  arXiv  Detail & Related papers  (2025-05-20T14:16:49Z)
• Distributionally Robust Models with Parametric Likelihood Ratios [123.05074253513935]
  Three simple ideas allow us to train models with DRO using a broader class of parametric likelihood ratios. We find that models trained with the resulting parametric adversaries are consistently more robust to subpopulation shifts when compared to other DRO approaches.
  arXiv  Detail & Related papers  (2022-04-13T12:43:12Z)

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.