Reinforcement learning, including reinforcement learning with verifiable rewards (RLVR), has emerged as a powerful approach for LLM post-training. Central to these approaches is the design of the importance sampling (IS) ratio used in off-policy policy-gradient estimation. Existing methods face a fundamental bias-variance dilemma: token-level IS ratios, as adopted by PPO (Schulman et al., 2017) and GRPO (Shao et al., 2024), introduce bias by ignoring prefix state distribution mismatch; full sequence ratios provide exact trajectory-level correction but suffer from high variance due to the multiplicative accumulation of per-token ratios, while GSPO (Zheng et al., 2025) improves numerical stability via length normalization at the cost of deviating from the exact full-sequence IS correction. In this work, we identify the cumulative token IS ratio, the product of per-token ratios up to position $t$, as a theoretically principled solution to this dilemma. We prove that, under the token-level policy-gradient formulation, this ratio provides an unbiased prefix correction for each token-level gradient term and has strictly lower variance than the full sequence ratio. Building on this insight, we propose CTPO (Cumulative Token Policy Optimization), which combines the cumulative token IS ratio with position-adaptive clipping that scales log-space clip bounds according to the natural $\sqrt{t}$ growth of the cumulative log-ratio. This yields more consistent regularization across token positions. We implement and evaluate CTPO in the tool-integrated reasoning setting on several challenging mathematical reasoning benchmarks, achieving the best average performance across both model scales compared with strong GRPO and GSPO baselines. Code will be available at https://github.com/horizon-llm/CTPO.
Beyond Truncation: Rethinking LLM Decoding as Ensemble Pruning
Institute of Statistics and Big Data, Renmin University of China（中国人民大学统计与大数据研究院）；Big Data and Responsible Artificial Intelligence for National Governance, Renmin University of China（中国人民大学大数据与国家治理研究院）；Center for Applied Statistics, Institute of Statistics and Big Data, Renmin University of China（中国人民大学统计与大数据研究院应用统计中心）
When 1+1 Backfires: Rethinking Relevance-Diversity Balancing for Sample-Aware Visual Token Reduction
State Key Laboratory of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所人工智能安全国家重点实验室）；University of Chinese Academy of Sciences（中国科学院大学）；Beijing Academy of Artificial Intelligence（北京智源人工智能研究院）
Rethinking the Understanding Ability across LLMs through Mutual Information
Department of Industrial and Systems Engineering（工业与系统工程系）；University of Houston（休斯敦大学）；Bakar Computational Health Sciences Institute（巴卡计算健康科学研究所）；UCSF
Rethinking Entropy Interventions in RLVR: An Entropy Change Perspective
State Key Laboratory of Blockchain and Data Security（区块链与数据安全国家重点实验室）；Zhejiang University（浙江大学）；Tencent（腾讯）；Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（杭州高新技术区（滨江）区块链与数据安全研究院）
FlowCut: Rethinking Redundancy via Information Flow for Efficient Vision-Language Models
School of Computer Science and Technology, Huazhong University of Science and Technology（华中科技大学计算机科学与技术学院）；Xiaohongshu Inc.（小红书公司）；Shanghai Jiao Tong University（上海交通大学）
Rethinking Token Reduction for Large Vision-Language Models
College of Computer Science and Technology, Zhejiang University（浙江大学计算机科学与技术学院）；State Key Laboratory of Blockchain and Data Security, Zhejiang University（浙江大学区块链与数据安全国家重点实验室）；Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（杭州高新技术区（滨江）区块链与数据安全研究院）；School of Software Technology, Zhejiang University（浙江大学软件学院）；National University of Singapore（新加坡国立大学）；Alibaba Group（阿里巴巴集团）
Seeing the Trees for the Forest: Rethinking Weakly-Supervised Medical Visual Grounding
Australian Institute for Machine Learning, University of Adelaide（澳大利亚机器学习研究所，阿德莱德大学）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；Macquarie University（麦考瑞大学）；Hanoi University of Science and Technology（河内科学技术大学）；University of Wollongong（沃林根大学）；Flinders University（弗林德斯大学）
TokenCast: Forecasting Token Consumption During LLM Agent Execution
Sun Yat-Sen University（中山大学）；Rensselaer Polytechnic Institute（伦斯勒理工学院）；Southeast University（东南大学）；Hong Kong University of Science and Technology（香港科技大学）
Rethinking Tokenized Graph Transformers for Node Classification
School of Computer Science and Technology, Huazhong University of Science and Technology（华中科技大学计算机科学与技术学院）；Hopcroft Center on Computing Science, Huazhong University of Science and Technology（华中科技大学霍普克罗夫特计算科学中心）；Cornell University（康奈尔大学）
Université de Montréal（蒙特利尔大学）；McGill University（麦吉尔大学）；Mila - Quebec AI Institute（Mila - 魁北克人工智能研究所）；University of Waterloo（滑铁卢大学）；The Chinese University of Hong Kong（香港中文大学）；Huawei Noah’s Ark Lab（华为诺亚方舟实验室）
TC-LLaVA: Rethinking the Transfer from Image to Video Understanding with Temporal Considerations
The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；Tencent PCG（腾讯平台与内容事业群）；The Hong Kong University of Science and Technology（香港科技大学）；Dalian University of Technology（大连理工大学）
Shanghai Jiao Tong University（上海交通大学）；South China University of Technology（南方科技大学）；Beihang University（北航）；Huawei Technologies Ltd（华为技术有限公司）；Shanghai AI Lab（上海人工智能实验室）；Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
Unmute the Patch Tokens: Rethinking Probing in Multi-Label Audio Classification
University of Kassel（卡塞尔大学）；Fraunhofer IEE（弗劳恩霍夫研究所）；Ghent University（根特大学）；ML and Systems Biology, MPI of Biochemistry（生物化学Max Planck研究所）；INRIA Montpellier（蒙彼利埃INRIA）
Decentralized Attention Fails Centralized Signals: Rethinking Transformers for Medical Time Series
Department of Biomedical Engineering, PolyU（PolyU 生物医学工程系）；School of Nursing, PolyU（PolyU 护理学院）；Yau Mathematical Sciences Center, Tsinghua University（清华大学尤太数学科学中心）
The dot product self-attention is known to be central and indispensable to state-of-the-art Transformer models. But is it really required? This paper investigates the true importance and contribution of the dot product-based self-attention mechanism on the performance of Transformer models. Via extensive experiments, we find that (1) random alignment matrices surprisingly perform quite competitively and (2) learning attention weights from token-token (query-key) interactions is useful but not that important after all. To this end, we propose \textsc{Synthesizer}, a model that learns synthetic …
MOPD-Router: Rethinking Teacher Routing in Multi-Teacher On-Policy Distillation
Shanghai Jiao Tong University（上海交通大学）；GAIR；Alibaba Group（阿里巴巴集团）；University of Science and Technology of China（中国科学技术大学）；Shanghai Innovation Institute（上海创新研究院）
