ICML 2026 研究揭示：监督微调过度训练会通过熵塌陷导致强化学习下模型输出排名反转，对齐策略需要警惕。
arXiv:2606.18487v1 Announce Type: new Abstract: The standard heuristic of selecting the SFT checkpoint with the highest pass@1 for GRPO can fail when …
只需一个token就能轻松骗过LLM评判者，揭示AI评估体系的安全软肋。
arXiv:2507.08794v3 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly trusted as automated judges, assisting evaluat…
新论文ReSum用强化学习协同LLM推理与摘要，解决长推理链低效问题，干货满满。
arXiv:2606.13316v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) is a central technique for improving long-horizo…
LLM推理训练新突破：通过训练时分解攻克零奖励难题，让模型从失败轨迹中学习！
arXiv:2606.09883v1 Announce Type: cross Abstract: Large language models (LLMs) have made remarkable progress in reasoning tasks, largely driven by pos…
提出专家评分标准解决RLVR中复杂约束问题，为强化学习奖励设计提供新范式
arXiv:2606.09118v1 Announce Type: new Abstract: As LLM capabilities advance rapidly, the evaluation methods used to assess them increasingly lag behin…
突破传统结果优化，动态跨模态协调强化学习提升视觉推理性能
arXiv:2606.08035v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a leading paradigm for enhancing …
CATPO方法通过批评增强的树策略优化，显著提升大语言模型推理中的密集奖励获取效率。
arXiv:2606.08346v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a dominant paradigm for improving t…
针对强化学习验证器数据需求难题，GeoMin用几何分布建模实现高效半监督学习，大幅降低标注成本
arXiv:2606.04516v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) significantly advances LLM reasoning, yet it f…
提出信用衰减特权反馈方法，引导搜索代理在硬问题上突破成功轨迹稀疏瓶颈
arXiv:2606.01830v1 Announce Type: new Abstract: Recent LLM search agents use reinforcement learning with verifiable rewards (RLVR) to learn search-aug…
提出GPOP方法，通过分组优先离轨优化提升LLM推理效率，有效剔除无效样本，为强化学习提供新思路。
arXiv:2606.01281v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for enhancing…
从因果机制出发，揭秘样本难度如何影响大模型强化学习中的推理表现
arXiv:2605.28388v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Reward (RLVR) is empirically shown to notably enhance the reaso…
提出冗余感知RLVR方法，改进多样本代码生成的评估指标
arXiv:2605.28022v1 Announce Type: new Abstract: LLMs for code generation are commonly evaluated in repeated-sampling settings using Pass@k, where mult…
从强化学习与虚拟现实融合的新视角切入，论文揭示时序调度比空间定位更关键，为RLVR优化提供突破性思路。
arXiv:2605.25381v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a core technique for post-training of…
揭秘大语言模型RLVR训练中的线性动力学机制，为强化学习优化提供新视角。
arXiv:2601.04537v3 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has driven significant performance gai…
锚定并强化视觉注意力，解决多模态大模型强化学习中的忠实感知难题
arXiv:2605.22072v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for advancin…
将组级别强化学习从奖励建模到策略优化统一为响应单纯形上的目标投影，提出Listwise Policy Optimization新方法。
arXiv:2605.06139v2 Announce Type: replace Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a standard approach for large lan…
新方法保障LLM在线部署每轮风险可控，基于共形预测与RLVR训练，安全认证更可靠。
arXiv:2605.20270v1 Announce Type: new Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on opera…
揭示RLVR训练中参数轨迹的秩一结构，仅需极小规模训练即可外推LLM推理能力，颠覆传统认知。
arXiv:2605.21468v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a dominant paradigm for improving rea…
提出逐步评分奖励机制，优化LLM推理的中间步骤监督，突破传统仅奖励最终答案的局限。
arXiv:2605.17291v1 Announce Type: new Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) is widely used to improve reasoning in large lan…
提出CuSearch课程采样法，通过搜索深度优化Agentic RAG的强化学习训练，提升效率
arXiv:2605.11611v2 Announce Type: replace Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a promising paradigm for trai…