DeepSeek（深度求索）

Correcting What You Cannot See: Credit Assignment for Perception Distillation in Multimodal Reasoners

修正不可见之缺陷：多模态推理器中感知蒸馏的信用分配

Feng Xiong, Leyan Xue, Hongyu Lin

机构 * Thinking Machines Lab（思维机器实验室） ； DeepSeek-AI（深度求索公司） 纠错

AI总结 该研究针对多模态推理器的感知蒸馏信用分配问题，提出感知修正蒸馏（PCD）方法，经8个基准测试，可提升不同规模模型的宏平均和结果，消融实验验证其有效性。

The Range Shrinks, the Threat Remains: Re-evaluating LLM Package Hallucinations on the 2026 Frontier-Model Cohort

范围缩小，威胁依旧：重新评估2026前沿模型队列上的LLM包幻觉

Aleksandr Churilov

机构 * Anthropic ； OpenAI ； Google ； DeepSeek 纠错

AI总结 本文重新评估了2026前沿模型队列上大型语言模型（LLM）的包幻觉现象，发现尽管幻觉率有所降低，但仍然存在威胁，识别出一组127个包名（109个在PyPI，18个在npm）被所有评估模型一致生成，构成一个跨模型的供应链攻击面，同时发现Python与JavaScript幻觉的不对称性以及DeepSeek V3.2和GPT-5.4-mini之间的高相似性。

Comments 13 pages, 3 figures, 4 tables. v2: incorporates coordinated-disclosure feedback from PyPI Security and Socket.dev; registrable attack surface refined to 53 names (41 PyPI, 12 npm). Headline rates unchanged. Replication of Spracklen et al. (USENIX Security 2025). Data and code: https://github.com/churik5/slopsquatting-replication-2026 and https://doi.org/10.5281/zenodo.19859120

Shieldstral

护盾斯特拉尔

Antonia Calvi, Avinash Sooriyarachchi, Giada Pistilli, Guillaume Lample, Maarten Buyl, Maximilian Augustin, Maximilian Müller, Pierre Stock, Tom Bewley, Wassim Bouaziz, Yimu Pan, Abdelaziz Bounhar, Abhijeet Somani, Aditi Kabra, Adrian Valente, Adrien Petralia, Adrien Sadé, Alan Jeffares, Albert Jiang, Aleksandr Timashov, Alexandre Cahill, Alexandre Gavaudan, Alexandre Laval, Alexandre Sablayrolles, Amélie Héliou, Amos You, André Jonasson, Andrew Bai, Andrew Ehrenberg, Andrew Zhao, Angele Lenglemetz, Anmol Agarwal, Arata Suzuki, Arjun Majumdar, Arthur Fournier, Artjom Joosen, Aylin Guliz Akkus, Aysenur Karaduman, Baptiste Bout, Baptiste Rozière, Baudouin De Monicault, Benjamin Holzschuh, Benjamin Lefaudeux, Benjamin Tibi, Bernhard Stadlbauer, Błażej Osiński, Camille Le Scao, Chaoran Yu, Charlotte Cronjäger, Chen-Yo Sun, Chris Bamford, Christian Wallenwein, Christophe Renaudin, Clémence Lanfranchi, Corentin Barreau, Corentin Sautier, Cristiana-Diana Diaconu, Cyprien Courtot, Daniel Marczak, Darius Dabert, Diego de Las Casas, Dominik Nuss, Dylan Rubini, Dzmitry Soupel, Elizaveta Demyanenko, Elliot Chane-Sane, Emilien Fugier, Emmanuel Gottlob, Erik Aas, Etienne Goffinet, Étienne Millon, Eujeong Choi, Fabian Paischer, Fabian Schlager, Faruk Ahmed, Federico Baldassarre, Filip Szatkowski, Florian Wiesner, Gabrielle Berrada, Gaëtan Ecrepont, Gaétan Lepage, Gaspard Blanchet, Gaspard Donada-Vidal, Gauthier Delerce, Gauthier Guinet, Genevieve Hayes, Georgii Novikov, Gianluca Galletti, Guillaume Breton, Guillaume Kunsch, Guillaume Martin, Guillaume Raille, Gunjan Dhanuka, Gunshi Gupta, Han Zhou, Harshil Shah, Hasan Furkan Vural, Hédi Hadiji, Hope McGovern, Hugo Cisneros, Hugo Thimonier, Indraneel Mukherjee, Ivan Cuevas Salazar, Jacques Sun, Jan Ludziejewski, Jason Rute, Jean Quentin, Jean-Hadrien Chabran, Jean-Malo Delignon, Jie Zhang, Joachim Studnia, Joep Barmentlo, Johannes Brandstetter, John Harvill, Jonas Amar, Jonas Schweizer, Joséphine Delas, Josselin Somerville, Julien Denize, Julien Tauran, Kartik Khandelwal, Khyathi Raghavi Chandu, Kilian Tep, Kush Jain, Larissa Laich, Laura Calem, Laurence Aitchison, Laurent Callot, Laurent Fainsin, Léo Cotteleer, Léonard Blier, Lingxiao Zhao, Louis Martin, Louis Serrano, Lucile Saulnier, Ludovic Ho Fuh, Luis Montero, Manon Chossegros, Marcin Możejko, Margaret Jennings, Markus Hennerbichler, Martin Alexandre, Mathieu Poirée, Mathieu Schmitt, Mathilde Guillaumin, Matthieu André, Matthieu Dinot, Matthieu Futeral, Maurits Bleeker, Mauro Comi, Max Mynter, Maxim Berman, Maxime Darrin, Maxime Louis, Melina Jingting Laimon, Mert Unsal, Mia Chiquier, Michael Pilcer, Michał Pietruszka, Michał Zając, Mikhail Biriuchinskii, Minh-Quang Pham, Minwoo Kang, Morgane Rivière, Namit Katariya, Nathan Grinsztajn, Nathan Simpson, Neeraj Aggarwal, Neha Gupta, Ola Mysiak, Oliver Leicht, Olivier Bousquet, Olivier Duchenne, Parag Jain, Patricia Wang, Patrick Blies, Patrick von Platen, Paul Jacob, Paul Wambergue, Paula Kurylowicz, Pavan Kumar Reddy, Pavel Kuksa, Philippe Pinel, Philomène Chagniot, Pierre-André Savalle, Piotr Milos, Prateek Gupta, Pravesh Agrawal, Quentin Desreumaux, Quentin Torroba, Quercus Hernandez, Ram Ramrakhya, Randall Isenhour, Ranjit Parva, Raul Perez Pelaez, Reinhard Sonnleitner, Rémi Delacourt, Richard Kurle, Rishi Shah, Rob Romijnders, Rohin Arora, Romain Sauvestre, Roman Soletskyi, Rosalie Millner, Rupert Menneer, Sagar Vaze, Samuel Barry, Samuel Belkadi, Samuel Humeau, Sanchit Gandhi, Sandeep Subramanian, Sarthak Mittal, Saskia Adaime, Sean Cha, Sebastian Kaltenbach, Shashwat Dalal, Shashwat Verma, Sherif Waly, Shrimai Prabhumoye, Siddhant Waghjale, Siddharth Gandhi, Simon Lepage, Simon Sorg, Soham Ghosh, Sophie Marbach, Srijan Mishra, Stanislas Lange, Steve Hong, Sumukh Aithal, Szymon Antoniak, Tarun Kumar Vangani, Teven Le Scao, Théo Cachet, Thibaut Lavril, Thomas Chabal, Thomas Coste, Thomas Defard, Thomas Foubert, Thomas Robert, Thomas Wang, Tianyu Zhang, Tim Lawson, Timothée Lacroix, Tobias Kronlachner, Tom Edwards, Tomas Hodan, Tuhin Das, Tyler Wang, Ulrick BLE, Umar Jamil, Umberto Tomasini, Valentin Macé, Van Phung, Vedant Nanda, Victor Jouault, Victor Letzelter, Victor Paltz, Victor Poucheret, Vincent Maladière, Vincent Pfister, Virgile Richard, Vladislav Bataev, Wen Ding Li, William Havard, William Marshall, Xinghui Li, Xingran Guo, Xinyu Yang, Yann Dreze, Yassine El Ouahidi, Yassir Bendou, Yihan Wang, Yves Martin des Taillades, Zaccharie Ramzi, Zhenlin Xu, Zsofia Csakany

机构 * OpenAI ； Anthropic ； Google DeepMind（谷歌DeepMind） ； DeepSeek-AI（深势科技人工智能） 纠错

AI总结 研究提出护盾斯特拉尔这一多模态安全分类器，将内容审核设为二元问答任务，通过特定数据构建方法及评估集，让小的自适应模型在文本安全基准测试中表现出色，在多模态安全分类上达新水平。

Variational-Ising-Attention (VIA):TailoredAttentionMattersfor Science

变分伊辛注意力（VIA）：适用于科学领域的定制注意力很重要

Rui Wang

机构 * DeepSeek-AI（深势科技人工智能） ； Moonshot AI（月之暗面人工智能） 纠错

AI总结 研究针对科学任务中softmax注意力独立性假设的不足，提出变分伊辛注意力（VIA），通过伊辛模型增强softmax归一化，在逆合成反应中心预测任务中实例化VIA，实验证明其优于标准softmax注意力，强调科学问题需定制注意力。

Comments 13 pages, ~20 figures

GuardianAgentBench: Where Agents Fail and How to Guard Them

GuardianAgentBench：智能体何处失败及如何防范

Vishal Ishwar Naik, Chenyu Xu, Donna Dong, Hussein Hassan, Abhishek Pradhan, Ofer Mendelevitch, Tallat Shafat, Humayun Irshad

机构 * Vectara, Inc.（Vectara公司） ； Anthropic（Anthropic公司） ； OpenAI（OpenAI公司） ； Google DeepMind（谷歌DeepMind） ； DeepSeek-AI（DeepSeek人工智能公司） 纠错

AI总结 研究大语言模型智能体安全可靠行为问题，提出GuardianAgentBench基准测试，含多阶段验证和攻击模式。实验发现模型有两种失败模式，性能随工具集和轮次深度下降。护栏实现优于系统提示防御，证明执行时结构干预可提升安全性且不影响正确行为。

Diagnosing Overhead in Dispatch Operations: Cross-architecture Observatory

调度操作中的开销诊断：跨架构观测站

Bole Ma, Jan Eitzinger, Harald Koestler, Gerhard Wellein

机构 * DeepSeek-V2-Lite MLA ； DeepSeek-MoE-16B MHA ； Qwen3-30B GQA ； Nemotron-30B Mamba-2 ； Qwen3.5-35B GDN 纠错

AI总结 该研究通过测试四个缓解方案的假设，发现扩展进程（EP）规模变化对专家最大/均值token比率的影响最多为5%，并且mock-token基准在路由Gini系数和批大小缩放趋势上存在高估。研究发现五种架构在相同矩阵中形成两个稳定的带状分布，这些带状分布而非EP度或mock数据配置是AlltoAll-aware互连和调度设计的正确工作负载输入。

Comments To be published in IEEE Hot Interconnects Symposium 2026

UMoE:Unlocking Every Expert in Domain-Specific Training

UMoE：在特定领域训练中解锁每个专家

Xuefeng Li, Pengfei Liu

机构 * Qwen Team（通义千问团队） ； DeepSeek-AI（渊亭科技） ； Thinking Machines Lab（思维机器实验室） ； LLM-Core Xiaomi（小米大语言模型核心团队） 纠错

AI总结 研究针对特定领域训练中专家池不合理问题，提出UMoE流程，先修剪低显著性专家，再扩展专家池，最后应用标准SFT，该方法在多架构、多领域及多基准测试中优于直接SFT，能转化冗余容量，提升训练效果。

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

通过可扩展查找实现条件记忆：大语言模型稀疏性的新轴

Xin Cheng, Rui Tian, Wangding Zeng, Damai Dai, Qinyu Chen, Bingxuan Wang, Zhenda Xie, Kezhao Huang, Xingkai Yu, Chengqi Deng, Shangyan Zhou, Chenggang Zhao, Zhewen Hao, Yukun Li, Han Zhang, Zhengyan Zhang, Yixu Wei, M. Y Xu, Huishuai Zhang, Dongyan Zhao, Wenfeng Liang

机构 * Peking University（北京大学） ； DeepSeek-AI 纠错

AI总结 研究针对Transformer缺乏知识查找原语的问题，引入条件记忆及Engram模块，通过制定稀疏分配问题发现U形缩放定律，扩展Engram至27B参数，在多领域提升性能，揭示其优势，为下一代稀疏模型提供重要建模原语。

DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation

DSpark：结合半自回归生成的置信度调度投机解码框架

Xin Cheng, Xingkai Yu, Chenze Shao, Jiashi Li, Yunfan Xiong, Yi Qian, Jiaqi Zhu, Shirong Ma, Xiaokang Zhang, Jiasheng Ye, Qinyu Chen, Chengqi Deng, Jiping Yu, Damai Dai, Zhengyan Zhang, Yixuan Wei, Yixuan Tan, Wenkai Yang, Runxin Xu, Yu Wu, Zhean Xu, Xuanyu Wang, Muyang Chen, Rui Tian, Xiao Bi, Zhewen Hao, Shaoyuan Chen, Huanqi Cao, Wentao Zhang, Anyi Xu, Huishuai Zhang, Dongyan Zhao, Wenfeng Liang

机构 * Peking University（北京大学） ； DeepSeek-AI（深势科技人工智能） 纠错

AI总结 针对现有投机解码接受率衰减、高并发下吞吐量劣化问题，DSpark采用半自回归架构与置信度调度验证，在实测中显著提升大模型推理速度与系统性能边界。

LionMuon: Alternating Spectral and Sign Descent for Efficient Training

LionMuon: 交替频谱和符号下降用于高效训练

Arman Bolatov, Artem Riabinin, Nikita Kornilov, Andrey Veprikov, Samuel Horváth, Martin Takáč, Aleksandr Beznosikov

机构 * DeepSeek-AI ； Essential AI ； Kimi Team 纠错

AI总结 本文提出LionMuon优化器，通过交替使用Lion和Muon的更新步骤，在保持有效性的同时显著降低平均迭代成本，同时证明了在重尾噪声下的复杂性界限，展示了其在不同模型规模下的优势。

Comments 38 pages, 13 figures, 4 tables

DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence

DeepSeek-V4: 迈向高效百万令牌上下文智能

DeepSeek-AI, Anyi Xu, Bangcai Lin, Bing Xue, Bingxuan Wang, Bingzheng Xu, Bochao Wu, Bowei Zhang, Chaofan Lin, Chen Dong, Chenchen Ling, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chengyu Hou, Chenhao Xu, Chenze Shao, Chong Ruan, Conner Sun, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Donghao Li, Dongjie Ji, Erhang Li, Fang Wei, Fangyun Lin, Fangzhou Yuan, Feiyu Xia, Fucong Dai, Guangbo Hao, Guanting Chen, Guoai Cao, Guolai Meng, Guowei Li, Han Yu, Han Zhang, Hanwei Xu, Hao Li, Haofen Liang, Haoling Zhang, Haoming Luo, Haoran Wei, Haotian Yuan, Haowei Zhang, Haowen Luo, Haoyu Chen, Haozhe Ji, Hengqing Zhang, Honghui Ding, Hongxuan Tang, Huanqi Cao, Huazuo Gao, Hui Qu, Hui Zeng, J Yang, JQ Zhu, Jia Luo, Jia Song, Jia Yu, Jialiang Huang, Jialu Cai, Jian Liang, Jiangting Zhou, Jiasheng Ye, Jiashi Li, Jiaxin Xu, Jiewen Hu, Jieyu Yang, Jin Chen, Jin Yan, Jingchang Chen, Jingli Zhou, Jingting Xiang, Jingyang Yuan, Jingyuan Cheng, Jingzi Zhou, Jinhua Zhu, Jiping Yu, Joseph Sun, Jun Ran, Junguang Jiang, Junjie Qiu, Junlong Li, Junmin Zheng, Junxiao Song, Kai Dong, Kaige Gao, Kang Guan, Kexing Zhou, Kezhao Huang, Kuai Yu, Lean Wang, Lecong Zhang, Lei Wang, Leyi Xia, Li Zhang, Liang Zhao, Lihua Guo, Lingxiao Luo, Linwang Ma, Linyan Zhu, Litong Wang, Liyu Cai, Liyue Zhang, Longhao Chen, MS Di, MY Xu, Max Mei, Miaojun Wang, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Mingming Li, Mingxu Zhou, Minmin Han, Ning Wang, Panpan Huang, Panpan Wang, Peixin Cong, Peiyi Wang, Peng Zhang, Qiancheng Wang, Qihao Zhu, Qingyang Li, Qinyu Chen, Qiushi Du, Qiwei Jiang, Rui Tian, Ruifan Xu, Ruijie Lu, Ruiling Xu, Ruiqi Ge, Ruisong Zhang, Ruizhe Pan, Runji Wang, Runqian Chen, Runqiu Yin, Runxin Xu, Ruomeng Shen, Ruoyu Zhang, Ruyi Chen, SH Liu, Shanghao Lu, Shangmian Sun, Shangyan Zhou, Shanhuang Chen, Shaofei Cai, Shaoheng Nie, Shaoqing Wu, Shaoyuan Chen, Shengding Hu, Shengyu Liu, Shiqiang Hu, Shirong Ma, Shiyu Wang, Shuiping Yu, Shunfeng Zhou, Shuting Pan, Shuying Yu, Songyang Zhou, Tao Ni, Tao Yun, Tian Jin, Tian Pei, Tian Ye, Tianle Lin, Tianran Ji, Tianyi Cui, Tianyuan Yue, Tingting Yu, Tun Wang, W Zhang, WL Xiao, Wangding Zeng, Wei An, Weilin Zhao, Wen Liu, Wenfeng Liang, Wenjie Pang, Wenjing Luo, Wenjing Yao, Wenjun Gao, Wenkai Yang, Wenlve Huang, Wenqing Hou, Wentao Zhang, Wenting Ma, Xi Gao, Xiang He, Xiangwen Wang, Xianzu Wang, Xiao Bi, Xiaodong Liu, Xiaohan Wang, Xiaokang Chen, Xiaokang Zhang, Xiaotao Nie, Xiaowen Sun, Xiaoxiang Wang, Xin Cheng, Xin Liu, Xin Xie, Xingchao Liu, Xingchen Liu, Xingkai Yu, Xingyou Li, Xinyu Yang, Xinyu Zhang, Xu Chen, Xuanyu Wang, Xuecheng Su, Xueyin Chen, Xuheng Lin, Xuwei Fu, YC Yan, YQ Wang, YW Ma, Yanfeng Luo, Yang Zhang, Yanhong Xu, Yanru Ma, Yanwen Huang, Yao Li, Yao Li, Yao Xu, Yao Zhao, Yaofeng Sun, Yaohui Wang, Yi Qian, Yi Shao, Yi Yu, Yichao Zhang, Yifan Ding, Yifan Shi, Yijia Wu, Yiliang Xiong, Yiling Ma, Ying He, Ying Tang, Ying Zhou, Yingjia Luo, Yinmin Zhong, Yishi Piao, Yisong Wang, Yixiang Zhang, Yixiao Chen, Yixuan Tan, Yixuan Wei, Yiyang Ma, Yiyuan Liu, Yonglun Yang, Yongqiang Guo, Yongtong Wu, Yu Wu, YuKun Li, Yuan Cheng, Yuan Ou, Yuanfan Xu, Yuanhao Li, Yuduan Wang, Yuehan Yang, Yuer Xu, Yuhan Wu, Yuhao Meng, Yuheng Zou, Yukun Zha, Yunfan Xiong, Yupeng Chen, Yuping Lin, Yuqian Cao, Yuqian Wang, Yushun Zhang, Yuting Yan, Yutong Lin, Yuxian Gu, Yuxiang Luo, Yuxiang You, Yuxuan Liu, Yuxuan Zhou, Yuyang Zhou, Yuzhen Huang, ZF Wu, Zehao Wang, Zehua Zhao, Zehui Ren, Zekai Zhang, Zhangli Sha, Zhe Fu, Zhe Ju, Zhean Xu, Zhenda Xie, Zhengyan Zhang, Zheren Gao, Zhewen Hao, Zhibin Gou, Zhicheng Ma, Zhigang Yan, Zhihong Shao, Zhixian Huang, Zhixuan Chen, Zhiyu Wu, Zhizhou Ren, Zhongyu Wu, Zhuoshu Li, Zhuping Zhang, Zian Xu, Zihao Wang, Zihua Qu, Zihui Gu, Zijia Zhu, Zilin Li, Zipeng Zhang, Ziwei Xie, Ziyi Gao, Ziyi Wan, Zizheng Pan, Zongqing Yao

机构 * DeepSeek-AI（深度求索人工智能） 纠错

AI总结 提出DeepSeek-V4系列MoE模型，通过混合注意力架构、流形约束超连接和Muon优化器，实现百万令牌上下文的高效推理，在核心任务上超越前代。

Reinforcement Learning for LLM-based Event Forecasting

基于强化学习的LLM事件预测

Amit Arnold Levy

机构 * Advanced Computer Science（高级计算机科学） ； DeepSeek R1 纠错

AI总结 使用GRPO微调LLM，结合Wikipedia修订工具获取实时信息，预测未来事件，使1.5B参数模型性能超越Claude Sonnet 3.5。

Comments Submitted internally at the University of Oxford in Oct 2025, migrated to arXiv on Jun 2026

MixRea: Benchmarking Explicit-Implicit Reasoning in Large Language Models

MixRea: 在大型语言模型中评估显式-隐式推理的基准测试

Yuanqing Cai, Ziyi Huang, Minhao Liu, Lixin Duan, Wen Li, Yanru Zhang

机构 * DeepMind ； DeepSeek-AI ； Anthropic ； Qwen ； Touvron ； Team 纠错

AI总结 本文提出MixRea基准测试，用于评估大型语言模型在显式和隐式推理任务中的表现，发现即使最佳模型也存在注意力不足的问题，并提出PRCP方法来改进推理能力。

Comments 12 pages, 6 figures, 4 tables

Scalable Knowledge Editing for Mixture-of-Experts LLMs via Tensor-Structured Updates

基于张量结构更新的混合专家LLM可扩展知识编辑

Roman Maksimov, Vladimir Aletov, Dmitry Bylinkin, Daniil Medyakov, Vladimir Solodkin, Aleksandr Beznosikov

机构 * OpenAI ； DeepSeek-AI ； Qwen Team（Qwen团队） ； Shazeer et al.（Shazeer等人） ； Molodtsov et al.（Molodtsov等人） 纠错

AI总结 本文提出一种针对混合专家架构LLM的知识编辑方法，通过张量结构和Woodbury矩阵恒等式实现高效参数更新，提升编辑效率6倍，扩展了知识编辑的应用范围。

Comments 17 pages, 3 architectures, 1 figure, 6 tables

Learning to Reason at the Frontier of Learnability

在可学习性前沿的学习推理

Thomas Foster, Anya Sims, Johannes Forkel, Mattie Fellows, Jakob Foerster

机构 * DeepSeek-R1 ； Tulu ； OpenAI 纠错

AI总结 本文提出通过可学习性采样方法优化大语言模型强化学习阶段，提升训练效率和效果，针对高变异性问题进行课程学习。

Pointer-CAD: Unifying B-Rep and Command Sequences via Pointer-based Edges & Faces Selection

Pointer-CAD：通过基于指针的边与面选择统一B-Rep和命令序列

Dacheng Qi, Chenyu Wang, Jingwei Xu, Tianzhe Chu, Zibo Zhao, Wen Liu, Wenrui Ding, Yi Ma, Shenghua Gao

机构 * Transcengram ； Beihang University（北京航空航天大学） ； The University of Hong Kong（香港大学） ； Shenzhen Loop Area Institute（深圳河套学院） ； ShanghaiTech University（上海科技大学） ； DeepSeek ； University of California, Berkeley（加州大学伯克利分校） 纠错

AI总结 Pointer-CAD通过基于指针的命令序列表示，将B-Rep模型的几何信息融入顺序建模，有效生成复杂几何结构并降低分割误差，显著提升CAD生成精度。

Comments Accepted by CVPR2026

Less Is More: Fast and Accurate Reasoning with Cross-Head Unified Sparse Attention

少即是多：基于交叉头统一稀疏注意力的快速准确推理

Lijie Yang, Zhihao Zhang, Arti Jain, Shijie Cao, Baihong Yuan, Yiwei Chen, Zhihao Jia, Ravi Netravali

机构 * DeepSeek-AI ； DeepMind ； OpenAI ； Team 纠错

AI总结 本文提出LessIsMore方法，通过交叉头统一稀疏注意力机制提升长 horizon 推理效率，减少 token 数量同时保持高精度，实现更快的端到端解码速度和稀疏注意力计算速度。

The Spectral Geometry of Thought: Phase Transitions, Instruction Reversal, Token-Level Dynamics, and Perfect Correctness Prediction in How Transformers Reason

思维的谱几何：相变、指令反向、令牌级动态以及完美正确性预测在Transformer推理中的表现

Yi Liu

机构 * DeepSeek-R1（深Seek-R1） ； Qwen（通义千问） ； Pythia ； Llama 纠错

AI总结 研究通过系统谱分析发现，大语言模型在推理与事实回忆时在隐藏激活空间中表现出谱相变现象，揭示了推理的谱压缩、指令调谐的谱反向等七种核心现象，建立了一套关于Transformer推理的谱理论。

DeepSeek-OCR 2: Visual Causal Flow

DeepSeek-OCR 2: 视觉因果流

Haoran Wei, Yaofeng Sun, Yukun Li

机构 * DeepSeek-AI 纠错

AI总结 DeepSeek-OCR 2通过两个级联的一维因果推理结构实现二维图像理解，提升视觉处理的语义连贯性与灵活性。

MHA2MLA-VLM: Enabling DeepSeek's Economical Multi-Head Latent Attention across Vision-Language Models

MHA2MLA-VLM: 使DeepSeek的经济型多头潜在注意力在视觉-语言模型中生效

Xiaoran Fan, Zhichao Sun, Tao Ji, Lixing Shen, Tao Gui

机构 * DeepSeek 纠错

AI总结 MHA2MLA-VLM通过参数高效的方法将现有视觉-语言模型转换为多头潜在注意力架构，减少缓存足迹并提升推理效率。

IIB-LPO: Latent Policy Optimization via Iterative Information Bottleneck

IIB-LPO: 通过迭代信息瓶颈进行潜在策略优化

Huilin Deng, Hongchen Luo, Yue Zhu, Long Li, Zhuoyue Chen, Xinghao Zhao, Ming Li, Jihai Zhang, Mengchang Wang, Yang Cao, Yu Kang

机构 * DeepSeek-AI 纠错

AI总结 IIB-LPO通过迭代信息瓶颈方法，解决LLM推理中的探索崩溃问题，实现更多样化的推理路径和更高的性能表现。

mHC: Manifold-Constrained Hyper-Connections

mHC：流形约束超连接

Zhenda Xie, Yixuan Wei, Huanqi Cao, Chenggang Zhao, Chengqi Deng, Jiashi Li, Damai Dai, Huazuo Gao, Jiang Chang, Kuai Yu, Liang Zhao, Shangyan Zhou, Zhean Xu, Zhengyan Zhang, Wangding Zeng, Shengding Hu, Yuqing Wang, Jingyang Yuan, Lean Wang, Wenfeng Liang

机构 * DeepSeek-AI（深Seek人工智能） 纠错

AI总结 mHC通过流形约束恢复残差连接的身份映射属性，提升模型的可扩展性和效率。

Insights into DeepSeek-V3: Scaling Challenges and Reflections on Hardware for AI Architectures

深入探讨DeepSeek-V3：扩展挑战与人工智能架构的硬件反思

Chenggang Zhao, Chengqi Deng, Chong Ruan, Damai Dai, Huazuo Gao, Jiashi Li, Liyue Zhang, Panpan Huang, Shangyan Zhou, Shirong Ma, Wenfeng Liang, Ying He, Yuqing Wang, Yuxuan Liu, Y. X. Wei

机构 * DeepSeek-AI（深Seek人工智能） 纠错

AI总结 DeepSeek-V3通过硬件感知的模型协同设计，解决了大规模LLM扩展中的硬件限制，展示了内存效率、计算优化和网络拓扑改进的创新方法。

Comments This is the author's version of the work. It is posted here for your personal use. Not for redistribution. The definitive version appeared as part of the Industry Track in Proceedings of the 52nd Annual International Symposium on Computer Architecture (ISCA '25)

Journal ref Proceedings of the 52nd Annual International Symposium on Computer Architecture (ISCA '25), June 2025

DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models

DeepSeek-V3.2: 推动开放大规模语言模型的前沿

DeepSeek-AI, Aixin Liu, Aoxue Mei, Bangcai Lin, Bing Xue, Bingxuan Wang, Bingzheng Xu, Bochao Wu, Bowei Zhang, Chaofan Lin, Chen Dong, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenhao Xu, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Erhang Li, Fangqi Zhou, Fangyun Lin, Fucong Dai, Guangbo Hao, Guanting Chen, Guowei Li, H. Zhang, Hanwei Xu, Hao Li, Haofen Liang, Haoran Wei, Haowei Zhang, Haowen Luo, Haozhe Ji, Honghui Ding, Hongxuan Tang, Huanqi Cao, Huazuo Gao, Hui Qu, Hui Zeng, Jialiang Huang, Jiashi Li, Jiaxin Xu, Jiewen Hu, Jingchang Chen, Jingting Xiang, Jingyang Yuan, Jingyuan Cheng, Jinhua Zhu, Jun Ran, Junguang Jiang, Junjie Qiu, Junlong Li, Junxiao Song, Kai Dong, Kaige Gao, Kang Guan, Kexin Huang, Kexing Zhou, Kezhao Huang, Kuai Yu, Lean Wang, Lecong Zhang, Lei Wang, Liang Zhao, Liangsheng Yin, Lihua Guo, Lingxiao Luo, Linwang Ma, Litong Wang, Liyue Zhang, M. S. Di, M. Y Xu, Mingchuan Zhang, Minghua Zhang, Minghui Tang, Mingxu Zhou, Panpan Huang, Peixin Cong, Peiyi Wang, Qiancheng Wang, Qihao Zhu, Qingyang Li, Qinyu Chen, Qiushi Du, Ruiling Xu, Ruiqi Ge, Ruisong Zhang, Ruizhe Pan, Runji Wang, Runqiu Yin, Runxin Xu, Ruomeng Shen, Ruoyu Zhang, S. H. Liu, Shanghao Lu, Shangyan Zhou, Shanhuang Chen, Shaofei Cai, Shaoyuan Chen, Shengding Hu, Shengyu Liu, Shiqiang Hu, Shirong Ma, Shiyu Wang, Shuiping Yu, Shunfeng Zhou, Shuting Pan, Songyang Zhou, Tao Ni, Tao Yun, Tian Pei, Tian Ye, Tianyuan Yue, Wangding Zeng, Wen Liu, Wenfeng Liang, Wenjie Pang, Wenjing Luo, Wenjun Gao, Wentao Zhang, Xi Gao, Xiangwen Wang, Xiao Bi, Xiaodong Liu, Xiaohan Wang, Xiaokang Chen, Xiaokang Zhang, Xiaotao Nie, Xin Cheng, Xin Liu, Xin Xie, Xingchao Liu, Xingkai Yu, Xingyou Li, Xinyu Yang, Xinyuan Li, Xu Chen, Xuecheng Su, Xuehai Pan, Xuheng Lin, Xuwei Fu, Y. Q. Wang, Yang Zhang, Yanhong Xu, Yanru Ma, Yao Li, Yao Li, Yao Zhao, Yaofeng Sun, Yaohui Wang, Yi Qian, Yi Yu, Yichao Zhang, Yifan Ding, Yifan Shi, Yiliang Xiong, Ying He, Ying Zhou, Yinmin Zhong, Yishi Piao, Yisong Wang, Yixiao Chen, Yixuan Tan, Yixuan Wei, Yiyang Ma, Yiyuan Liu, Yonglun Yang, Yongqiang Guo, Yongtong Wu, Yu Wu, Yuan Cheng, Yuan Ou, Yuanfan Xu, Yuduan Wang, Yue Gong, Yuhan Wu, Yuheng Zou, Yukun Li, Yunfan Xiong, Yuxiang Luo, Yuxiang You, Yuxuan Liu, Yuyang Zhou, Z. F. Wu, Z. Z. Ren, Zehua Zhao, Zehui Ren, Zhangli Sha, Zhe Fu, Zhean Xu, Zhenda Xie, Zhengyan Zhang, Zhewen Hao, Zhibin Gou, Zhicheng Ma, Zhigang Yan, Zhihong Shao, Zhixian Huang, Zhiyu Wu, Zhuoshu Li, Zhuping Zhang, Zian Xu, Zihao Wang, Zihui Gu, Zijia Zhu, Zilin Li, Zipeng Zhang, Ziwei Xie, Ziyi Gao, Zizheng Pan, Zongqing Yao, Bei Feng, Hui Li, J. L. Cai, Jiaqi Ni, Lei Xu, Meng Li, Ning Tian, R. J. Chen, R. L. Jin, S. S. Li, Shuang Zhou, Tianyu Sun, X. Q. Li, Xiangyue Jin, Xiaojin Shen, Xiaosha Chen, Xinnan Song, Xinyi Zhou, Y. X. Zhu, Yanping Huang, Yaohui Li, Yi Zheng, Yuchen Zhu, Yunxian Ma, Zhen Huang, Zhipeng Xu, Zhongyu Zhang, Dongjie Ji, Jian Liang, Jianzhong Guo, Jin Chen, Leyi Xia, Miaojun Wang, Mingming Li, Peng Zhang, Ruyi Chen, Shangmian Sun, Shaoqing Wu, Shengfeng Ye, T. Wang, W. L. Xiao, Wei An, Xianzu Wang, Xiaowen Sun, Xiaoxiang Wang, Ying Tang, Yukun Zha, Zekai Zhang, Zhe Ju, Zhen Zhang, Zihua Qu

机构 * DeepSeek-AI 纠错

AI总结 DeepSeek-V3.2通过高效注意力机制、强化学习框架和大规模代理任务合成流水线，在计算效率与推理能力上取得突破，超越GPT-5并获国际竞赛金牌。

DeepSeekMath-V2: Towards Self-Verifiable Mathematical Reasoning

DeepSeekMath-V2：迈向自我验证的数学推理

Zhihong Shao, Yuxiang Luo, Chengda Lu, Z. Z. Ren, Jiewen Hu, Tian Ye, Zhibin Gou, Shirong Ma, Xiaokang Zhang

机构 * DeepSeek-AI 纠错

AI总结 DeepSeekMath-V2通过自我验证机制提升数学推理能力，实现定理证明的高准确率和严谨性。