语言大模型 / LLM

Deadline-Driven Hierarchical Agentic Resource Sharing for AI Services and RAN Functions in AI-RAN

面向截止期限的分层代理资源共享框架用于AI服务和RAN功能在AI-RAN中

Haiyuan Li, Yulei Wu, Dimitra Simeonidou

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出了一种分层代理框架，结合大语言模型用于慢速时间尺度的AI服务和RAN功能放置，以及基于截止期限的凸优化算法用于快速时间尺度的GPU/CPU分配，以提高AI-RAN的资源利用率和SLA满足率。

When Eavesdroppers Reason: Agentic Eavesdropping Attacks on Semantic Communication

当窃听者进行推理：语义通信中的代理窃听攻击

Shunpu Tang, Qianqian Yang, Zhiguo Shi, Jiming Chen, Xuemin Shen

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出一种基于大语言模型的代理窃听者，通过闭环工作流中的三个功能代理，无需窃听信道状态信息即可实现超过75%的窃听成功率，揭示了语义通信中严重的隐私威胁。

Towards Compute-Aware In-Switch Computing for LLMs Tensor-Parallelism on Multi-GPU Systems

面向多GPU系统的计算感知交换机计算

Chen Zhang, Qijun Zhang, Zhuoshan Zhou, Yijia Diao, Haibo Wang, Zhe Zhou, Zhipeng Tu, Zhiyao Li, Guangyu Sun, Zhuoran Song, Zhigang Ji, Jingwen Leng, Minyi Guo

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出CAIS框架，通过计算感知的ISA扩展、线程块协调和数据流优化，提升多GPU系统中LLM张量并行的通信与计算效率，实验显示其在训练速度上优于现有方案。

Comments 15 pages, 18 figures, HPCA 2026

DEPTEX: Organization-First, Open Source Dependency Risk Monitoring

DEPTEX: 以组织为中心，开源依赖风险监控

Henry Ruckman-Utting, Vrushal Nedungadi, Taiga Okuma, LeTian Wang, Stephen Ehebald, Mohammad A. Tayebi

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 DEPTEX通过图谱技术与大语言模型结合，解决开源依赖风险评估中的语义缺失问题，实现动态合规管理和主动风险控制。

BLINC: Context-Specific Causal Learning for Automated RAN Configuration

BLINC：基于上下文的因果学习用于自动无线接入网配置

Reshma Prasad, Michele Polese, Tommaso Melodia

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 BLINC通过整合电信领域知识，利用贝叶斯网络框架实现无线接入网配置的自动优化，提升吞吐量63.5%，降低误块率19.7%。

Comments 10 pages

Generalizing Test Cases for Comprehensive Test Scenario Coverage

通用测试用例以实现全面的测试场景覆盖

Binhang Qi, Yun Lin, Xinyi Weng, Chenyan Liu, Hailong Sun, Gordon Fraser, Jin Song Dong

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出TestGeneralizer框架，通过增强需求理解、生成测试场景模板并生成可执行测试用例，提升测试场景覆盖。在12个开源Java项目上测试，相比ChatTester在突变和LLM评估中分别提升31.66%和23.08%。

Comments Accepted at FSE 2026

TaskLens: Generating Task-Conditioned Scaffolded Interfaces for Learning Professional Creative Software

TaskLens: 为学习专业创意软件生成任务条件的引导界面

Yimeng Liu, Misha Sra

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 TaskLens通过LLM生成任务条件的引导界面，降低新手学习难度，提升任务表现和领域概念学习。

Comments DIS '26 | ACM Designing Interactive Systems Conference

System-aware contextual digital twin for ICS anomaly diagnosis

面向系统的上下文数字孪生用于ICS异常诊断

Eungyu Woo, Yooshin Kim, Wonje Heo, Donghoon Shin

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出一种面向系统的无监督框架，结合轻量级在线检测与上下文解释，实现ICS异常诊断的实时检测和可解释诊断。

Where Relevance Emerges: A Layer-Wise Study of Internal Attention for Zero-Shot Re-Ranking

相关性如何出现：对零样本重排序内部注意力的分层研究

Haodong Chen, Shengyao Zhuang, Zheng Yao, Guido Zuccon, Teerapong Leelanupab

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文研究了多排名框架中生成、似然和内部注意力机制的对比，发现Transformer层中相关性信号呈现钟形分布，提出Selective-ICR策略降低推理延迟30%-50%，并在BRIGHT基准测试中证明了内部信号在复杂推理排序中的潜力。

Comments Accepted by SIGIR 2026. 10 pages, 5 figures, 4 tables. Code available at https://github.com/ielab/Selective-ICR

FlyCatcher: Neural Inference of Runtime Checkers from Tests

FlyCatcher: 从测试中自动推断运行时检查器

Beatriz Souza, Chang Lou, Suman Nath, Michael Pradel

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 FlyCatcher通过结合LLM合成、静态分析和动态验证，从现有测试中自动推断运行时检查器，能检测更多隐性故障。

iPDB -- Optimizing Semantic SQL Queries

iPDB -- 优化语义SQL查询

Udesh Kumarasinghe, Tyler Liu, Ahmed R. Mahmood, Chunwei Liu, Walid G. Aref

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 iPDB通过引入语义查询优化技术，支持在数据库内进行机器学习和大语言模型推理，实现语义SQL查询的高效执行，平均提速2.5倍，最高可达30倍。

Prefill-as-a-Service: KVCache of Next-Generation Models Could Go Cross-Datacenter

预填充即服务：下一代模型的KV缓存可跨数据中心

Ruoyu Qin, Weiran He, Yaoyu Wang, Zheming Li, Xinran Xu, Yongwei Wu, Weimin Zheng, Mingxing Zhang

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出PrfaaS架构，通过跨数据中心的预填充服务优化大规模LLM服务，实现异构部署和资源弹性，提升吞吐量和响应速度。

Comments 16 pages, 5 figures, 6 tables

Open-Architecture End-to-End System for Real-World Autonomous Robot Navigation

面向真实世界的自主机器人导航开架系统

Venkata Naren Devarakonda, Ali Umut Kaypak, Raktim Gautam Goswami, Naman Patel, Rooholla Khorrambakht, Prashanth Krishnamurthy, Farshad Khorrami

机构 * Control/Robotics Research Laboratory (CRRL), Department of Electrical and Computer Engineering, NYU Tandon School of Engineering（控制/机器人研究实验室（CRRL），电气与计算机工程系，NYU塔恩顿工程学院） 纠错

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出一种轻量级开架端到端系统，通过整合多组件实现四足机器人实时导航，利用自然语言任务和语义信息构建场景图，基于LLM生成动态计划，实现在多个室内环境中的零样本自主导航。

Structural Verification for Reliable EDA Code Generation without Tool-in-the-Loop Debugging

结构验证用于无需工具在回路调试的可靠EDA代码生成

Dinithi Jayasuriya, Aravind Saravanan, Nilesh Ahuja, Amanda Rios, Amit Trivedi

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出通过强制执行结构正确性来提高EDA代码生成的可靠性与效率，采用结构依赖图作为显式执行合同，并通过验证引导合成框架进行图条件检索、约束生成和分阶段预执行验证，从而在单步和多步任务中显著提升通过率。

HadAgent: Harness-Aware Decentralized Agentic AI Serving with Proof-of-Inference Blockchain Consensus

HadAgent：基于证明-推理区块链共识的去中心化代理AI服务

Landy Jimenez, Mariah Weatherspoon, Bingyu Shen, Yi Sheng, Jianming Liu, Boyang Li

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 HadAgent通过引入证明-推理共识机制，替代传统工作量证明，实现去中心化代理AI服务，提升验证效率与安全性，实验显示高检测率与低误报率。

Comments 9 pages, 5 figures

CIMple: Standard-cell SRAM-based CIM with LUT-based split softmax for attention acceleration

CIMple: 基于标准单元SRAM的CIM及其基于LUT的拆分softmax用于注意力加速

Bas Ahn, Xingjian Tao, Manil Dev Gomony, Marc Geilen, Henk Corporaal

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出CIMple，一种基于标准单元SRAM的CIM架构，通过8位并行权重喂送和LUT固定点实现，实现高效注意力加速，评估了28nm实现的32kb自注意力加速器性能。

Comments 10 pages, 11 figures

Journal ref 2025 Cross-Disciplinary Conference on Memory-Centric Computing (CCMCC), 1-10

BioVLM: Routing Prompts, Not Parameters, for Cross-Modality Generalization in Biomedical VLMs

BioVLM：通过路由提示而非参数实现生物医学VLMs的跨模态泛化

Mainak Singha, Tanisha Gupta, Ankit Jha, Muhammad Haris Khan, Sayantani Ghosh, Biplab Banerjee

机构 * University of Trento（特伦托大学） ； Carnegie Mellon University（卡内基梅隆大学） ； LNMIIT Jaipur（贾伊普尔 LNMIIT） ； MBZUAI ； Sunandan Divatia School of Science（Sunandan Divatia 科学学院） ； IIT Bombay（博伊斯大学） 纠错

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；language model(abstract)；pretraining(abstract)

AI总结 BioVLM通过动态提示选择和提示银行学习，提升跨域泛化能力，无需大量骨干网络微调，在11个MedMNIST+2D数据集上取得新突破。

Comments Accepted in ACL Findings 2026

TestDecision: Sequential Test Suite Generation via Greedy Optimization and Reinforcement Learning

TestDecision: 通过贪婪优化和强化学习进行序列测试套件生成

Guoqing Wang, Chengran Yang, Xiaoxuan Zhou, Zeyu Sun, Bo Wang, David Lo, Dan Hao

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出TestDecision，通过将LLM转化为神经贪婪专家，利用贪婪策略和强化学习生成高质量测试套件，显著提升分支覆盖率和执行通过率。

Comments 22 pages, 4 figures; corrected metadata; marked corresponding author

AI-assisted Protocol Information Extraction For Improved Accuracy and Efficiency in Clinical Trial Workflows

人工智能辅助的协议信息提取以提高临床试验工作流程的准确性和效率

Ramtin Babaeipour, François Charest, Madison Wright

机构 * Banting Health AI（巴丁健康AI） 纠错

专题命中 效率与部署 ：LLM(summary_cn,abstract_cn)；分类 cs.CL、cs.AI、cs.LG

AI总结 本文研究了利用AI系统进行临床试验协议信息提取，通过比较RAG方法与传统LLM的准确性，发现AI能显著提升提取效率和准确性，降低工作负担。

Comments Updated to accepted manuscript. Published in Journal of Biomedical Informatics, Volume 179, July 2026, 105036

Journal ref Journal of Biomedical Informatics, Volume 179, July 2026, 105036

One Token per Highly Selective Frame: Towards Extreme Compression for Long Video Understanding

每个高度选择性帧一个标记：朝着长视频理解的极端压缩

Zheyu Zhang, Ziqi Pang, Shixing Chen, Xiang Hao, Vimal Bhat, Yu-Xiong Wang

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Amazon Prime Video（亚马逊Prime视频） 纠错

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；large language model(abstract)；language model(abstract)

AI总结 本文提出XComp模型，通过极端视频标记压缩提升长视频理解性能，结合标记级和帧级压缩，实现更高的压缩比和更密集的帧采样。

Comments Appear in the proceedings of NeurIPS 2025

SWE-TRACE: Optimizing Long-Horizon SWE Agents Through Rubric Process Reward Models and Heuristic Test-Time Scaling

SWE-TRACE：通过规则过程奖励模型和启发式测试时间缩放优化长 horizon SWE代理

Hao Han, Jin Xie, Xuehao Ma, Weiquan Zhu, Ziyao Zhang, ZhiLiang Long, Hongkai Chen, Qingwen Ye

专题命中 效率与部署 ：LLM(abstract,abstract_cn)；SFT(abstract,abstract_cn)

AI总结 本文提出SWE-TRACE框架，通过优化数据收集、强化学习和测试时间推理，提升SWE代理的长horizon推理能力，减少token消耗和推理延迟。

Multi-Agent Object Detection Framework Based on Raspberry Pi YOLO Detector and Slack-Ollama Natural Language Interface

基于Raspberry Pi YOLO检测器和Slack-Ollama自然语言接口的多智能体目标检测框架

Vladimir Kalušev, Branko Brkljač, Milan Brkljač

机构 * The Institute for Artificial Intelligence Research（人工智能研究机构） ； Development of Serbia（塞尔维亚发展） ； Department of Power, Electronic and Telecommunication Engineering（电力、电子与电信工程系） ； Faculty of Technical Sciences, University of Novi Sad（技术科学学院，诺维萨德大学） ； Faculty of Finance, Banking and Auditing, Alfa BK University（金融、银行与审计学院，阿尔法BK大学） 纠错

专题命中 效率与部署 ：LLM(summary_cn,abstract)

AI总结 本文提出了一种基于边缘计算的多智能体目标检测系统，利用LLM自然语言接口实现系统控制与通信，并在资源受限硬件平台上集成所有组件。

Comments 19 pages, 7 figures, 2 tables, implementation code will be made available upon manuscript publication

Nemotron 3 Super: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning

Nemotron 3 Super：开放、高效的混合专家混合Mamba-Transformer模型用于代理推理

NVIDIA, :, Aakshita Chandiramani, Aaron Blakeman, Abdullahi Olaoye, Abhibha Gupta, Abhilash Somasamudramath, Abhinav Khattar, Adeola Adesoba, Adi Renduchintala, Adil Asif, Aditya Agrawal, Aditya Vavre, Ahmad Kiswani, Aishwarya Padmakumar, Ajay Hotchandani, Akanksha Shukla, Akhiad Bercovich, Aleksander Ficek, Aleksandr Shaposhnikov, Alex Gronskiy, Alex Kondratenko, Alex Neefus, Alex Steiner, Alex Yang, Alexander Bukharin, Alexander Young, Ali Hatamizadeh, Ali Taghibakhshi, Alina Galiautdinova, Alisa Liu, Alok Kumar, Ameya Sunil Mahabaleshwarkar, Amir Klein, Amit Zuker, Amnon Geifman, Anahita Bhiwandiwalla, Ananth Subramaniam, Andrew Tao, Anjaney Shrivastava, Anjulie Agrusa, Ankur Srivastava, Ankur Verma, Ann Guan, Anna Shors, Annamalai Chockalingam, Anubhav Mandarwal, Aparnaa Ramani, Arham Mehta, Arti Jain, Arun Venkatesan, Asha Anoosheh, Ashwath Aithal, Ashwin Poojary, Asif Ahamed, Asit Mishra, Asli Sabanci Demiroz, Asma Kuriparambil Thekkumpate, Atefeh Sohrabizadeh, Avinash Kaur, Ayush Dattagupta, Barath Subramaniam Anandan, Bardiya Sadeghi, Barnaby Simkin, Ben Lanir, Benedikt Schifferer, Benjamin Chislett, Besmira Nushi, Bilal Kartal, Bill Thiede, Bita Darvish Rouhani, Bobby Chen, Boris Ginsburg, Brandon Norick, Branislav Kisacanin, Brian Yu, Bryan Catanzaro, Buvaneswari Mani, Carlo del Mundo, Chankyu Lee, Chanran Kim, Chantal Hwang, Chao Ni, Charles Wang, Charlie Truong, Cheng-Ping Hsieh, Chenhan Yu, Chenjie Luo, Cherie Wang, Chetan Mungekar, Chintan Patel, Chris Alexiuk, Chris Holguin, Chris Wing, Christian Munley, Christopher Parisien, Chuck Desai, Chunyang Sheng, Collin Neale, Cyril Meurillon, Dakshi Kumar, Dan Gil, Dan Su, Dane Corneil, Daniel Afrimi, Daniel Burkhardt Eliuth Triana, Daniel Egert, Daniel Fatade, Daniel Lo, Daniel Rohrer, Daniel Serebrenik, Daniil Sorokin, Daria Gitman, Daria Levy, Darko Stosic, David Edelsohn, David Messina, David Mosallanezhad, David Tamok, Deena Donia, Deepak Narayanan, Devin O'Kelly, Dheeraj Peri, Dhruv Nathawani, Di Wu, Dima Rekesh, Dina Yared, Divyanshu Kakwani, Dmitry Konyagin Brandon Tuttle, Dong Ahn, Dongfu Jiang, Dorrin Poorkay, Douglas O'Flaherty, Duncan Riach, Dusan Stosic, Dustin Van Stee, Edgar Minasyan, Edward Lin, Eileen Peters Long, Elad Segal, Elena Lantz, Elena Lewis, Ellie Evans, Elliott Ning, Eric Chung, Eric Harper, Eric Pham-Hung, Eric W. Tramel, Erick Galinkin, Erik Pounds, Esti Etrog, Evan Briones, Evan Wu, Evelina Bakhturina, Evgeny Tsykunov, Ewa Dobrowolska, Farshad Saberi Movahed, Farzan Memarian, Fay Wang, Fei Jia, Felipe Soares, Felipe Vieira Frujeri, Feng Chen, Fengguang Lin, Ferenc Galko, Fortuna Zhang, Frankie Siino, Frida Hou, Gantavya Bhatt, Gargi Prasad, Geethapriya Venkataramani, Geetika Gupta, George Armstrong, Gerald Shen, Giulio Borghesi, Gordana Neskovic, Gorkem Batmaz, Grace Lam, Grace Wu, Greg Pauloski, Greyson Davis, Grigor Nalbandyan, Guoming Zhang, Guy Farber, Guyue Huang, Haifeng Qian, Haran Kumar Shiv Kumar, Harry Kim, Harsh Sharma, Hayate Iso, Hayley Ross, Herbert Hum, Herman Sahota, Hexin Wang, Himanshu Soni, Hiren Upadhyay, Huy Nguyen, Iain Cunningham, Ido Galil, Ido Shahaf, Igino Padovani, Igor Gitman, Igor Shovkun, Ikroop Dhillon, Ilya Loshchilov, Ingrid Kelly, Itamar Schen, Itay Levy, Ivan Moshkov, Izik Golan, Izzy Putterman, Jain Tu, Jan Baczek, Jan Kautz, Jane Polak Scowcroft, Janica Rosenberg, Jared Casper, Jarrod Pflum, Jason Grant, Jason Sewall, Jatin Mitra, Jeffrey Glick, Jenny Chen, Jesse Oliver, Jiacheng Xu, Jiafan Zhu, Jialin Song, Jian Zhang, Jiaqi Zeng, Jie Lou, Jill Milton, Jim Chow, Jimmy Zhang, Jinhang Choi, Jining Huang, Jocelyn Huang, Joel Caruso, Joey Conway, Joey Guman, Johan Jatko, John Kamalu, Johnny Greco, Jonathan Cohen, Jonathan Raiman, Joseph Jennings, Joyjit Daw, Juan Yu, Julio Tapia, Junkeun Yi, Jupinder Parmar, Jyothi Achar, Kari Briski, Kartik Mattoo, Katherine Cheung, Katherine Luna, Keith Wyss, Kevin Shih, Kezhi Kong, Khanh Nguyen, Khushi Bhardwaj, Kirill Buryak, Kirthi Shankar Sivamani, Konstantinos Krommydas, Kris Murphy, Krishna C. Puvvada, Krzysztof Pawelec, Kumar Anik, Laikh Tewari, Laya Sleiman, Leo Du, Leon Derczynski, Li Ding, Lilach Ilan, Lingjie Wu, Lizzie Wei, Luis Vega, Lun Su, Maarten Van Segbroeck, Maer Rodrigues de Melo, Magaret Zhang, Mahan Fathi, Makesh Narsimhan Sreedhar, Makesh Sreedhar, Makesh Tarun Chandran, Manuel Reyes Gomez, Maor Ashkenazi, Marc Cuevas, Marc Romeijn, Margaret Zhang, Mark Cai, Mark Gabel, Markus Kliegl, Martyna Patelka, Maryam Moosaei, Matthew Varacalli, Matvei Novikov, Mauricio Ferrato, Mehrzad Samadi, Melissa Corpuz, Meng Xin, Mengdi Wang, Mengru Wang, Meredith Price, Micah Schaffer, Michael Andersch, Michael Boone, Michael Evans, Michael Z Wang, Miguel Martinez, Mikail Khona, Mike Chrzanowski, Mike Hollinger, Mingyuan Ma, Minseok Lee, Mohammad Dabbah, Mohammad Shoeybi, Mostofa Patwary, Nabin Mulepati, Nader Khalil, Najeeb Nabwani, Nancy Agarwal, Nanthini Balasubramaniam, Narimane Hennouni, Narsi Kodukula, Natalie Hereth, Nathaniel Pinckney, Nave Assaf, Negar Habibi, Nestor Qin, Neta Zmora, Netanel Haber, Nick Reamaroon, Nickson Quak, Nidhi Bhatia, Nikhil Jukar, Nikki Pope, Nikolai Ludwig, Nima Tajbakhsh, Nir Ailon, Nirmal Juluru, Nirmalya De, Nowel Pitt, Oleg Rybakov, Oleksii Hrinchuk, Oleksii Kuchaiev, Olivier Delalleau, Oluwatobi Olabiyi, Omer Ullman Argov, Omri Almog, Omri Puny, Oren Tropp, Otavio Padovani, Ouye Xie, Parth Chadha, Pasha Shamis, Paul Gibbons, Pavlo Molchanov, Peter Belcak, Peter Jin, Pinky Xu, Piotr Januszewski, Pooya Jannaty, Prachi Shevate, Pradeep Thalasta, Pranav Prashant Thombre, Prasoon Varshney, Prerana Gambhir, Pritam Gundecha, Przemek Tredak, Qing Miao, Qiyu Wan, Quan Tran Minh, Rabeeh Karimi Mahabadi, Rachel Oberman, Rachit Garg, Rahul Kandu, Raina Zhong, Ran El-Yaniv, Ran Zilberstein, Rasoul Shafipour, Renee Yao, Renjie Pi, Richard Mazzarese, Richard Wang, Rick Izzo, Ridhima Singla, Rima Shahbazyan, Rishabh Garg, Ritika Borkar, Ritu Gala, Riyad Islam, Robert Clark, Robert Hesse, Roger Waleffe, Rohit Varma Kalidindi, Rohit Watve, Roi Koren, Ron Fan, Ruchika Kharwar, Ruisi Cai, Ruoxi Zhang, Russell J. Hewett, Ryan Prenger, Ryan Timbrook, Ryota Egashira, Sadegh Mahdavi, Sagar Singh Ashutosh Joshi, Sahil Modi, Samuel Kriman, Sandeep Pombra, Sanjay Kariyappa, Sanjeev Satheesh, Santiago Pombo, Saori Kaji, Satish Pasumarthi, Saurav Mishra, Saurav Muralidharan, Scott Hara, Sean Narenthiran, Sebastian Rogawski, Seonjin Na, Seonmyeong Bak, Sepehr Sameni, Seth Poulos, Shahar Mor, Shantanu Acharya, Shaona Ghosh Adam Lord, Sharath Turuvekere Sreenivas, Shaun Kotek, Shaya Gharghabi, Shelby Thomas, Sheng-Chieh Lin, Shibani Likhite, Shiqing Fan, Shiyang Chen, Shreya Gopal, Shrimai Prabhumoye, Shubham Pachori, Shubham Toshniwal, Shuo Zhang, Shuoyang Ding, Shyam Renjith, Shyamala Prayaga, Siddhartha Jain, Simeng Sun, Sirisha Rella, Sirshak Das, Smita Ithape, Sneha Harishchandra S, Somshubra Majumdar, Soumye Singhal, Sri Harsha Singudasu, Sriharsha Niverty, Stas Sergienko, Stefana Gloginic, Stefania Alborghetti, Stephen Ge, Stephen McCullough, Sugam Dipak Devare, Suguna Varshini Velury, Sukrit Rao, Sumeet Kumar Barua, Sunny Gai, Suseella Panguluri, Sushil Koundinyan, Swathi Patnam, Sweta Priyadarshi, Swetha Bhendigeri, Syeda Nahida Akter, Sylendran Arunagiri, Tailling Yuan, Talor Abramovich, Tan Bui, Tan Yu, Terry Kong, Thanh Do, Thomas Gburek, Thorgane Marques, Tiffany Moore, Tijmen Blankevoort, Tim Moon, Timothy Ma, Tiyasa Mitra, Tomasz Grzegorzek, Tomer Asida, Tomer Bar Natan, Tomer Keren, Tomer Ronen, Traian Rebedea, Trenton Starkey, Tugrul Konuk, Twinkle Vashishth, Tyler Condensa, Udi Karpas, Ushnish De, Vahid Noorozi, Vahid Noroozi, Vanshil Atul Shah, Veena Vaidyanathan, Venkat Srinivasan, Venmugil Elango, Victor Cui, Vijay Korthikanti, Vikas Mehta, Virginia Adams, Virginia Wu, Vitaly Kurin, Vitaly Lavrukhin, Vladimir Anisimov, Wan Seo, Wanli Jiang, Wasi Uddin Ahmad, Wei Du, Wei Ping, Wei-Ming Chen, Wendy Quan, Wenliang Dai, Wenwen Gao, Will Jennings, William Zhang, Xiaowei Ren, Xiaowen Xin, Xin Li, Yang Yu, Yangyi Chen, Yaniv Galron, Yashaswi Karnati, Yejin Choi, Yev Meyer, Yi-Fu Wu, Yian Zhang, Ying Lin, Yonatan Geifman, Yonggan Fu, Yoshi Suhara, Youngeun Kwon, Yuan Zhang, Yuki Huang, Zach Moshe, Zhilin Wang, Zhiyu Cheng, Zhongbo Zhu, Zhuolin Yang, Zihan Liu, Zijia Chen, Zijie Yan, Zuhair Ahmed

机构 * NVIDIA 纠错

专题命中 效率与部署 ：SFT(abstract,abstract_cn)；post-training(abstract)；分类 cs.CL、cs.AI、cs.LG

AI总结 Nemotron 3 Super是首个采用NVFP4预训练、LatentMoE架构和MTP层加速推理的混合Mamba-Transformer模型，实现更高的推理吞吐量和准确率。

SWE-Shepherd: Advancing PRMs for Reinforcing Code Agents

SWE-Shepherd：推进基于强化学习的代码代理的PRMs

Mahir Labib Dihan, Md Ashrafur Rahman Khan

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；prompting(abstract)

AI总结 本文提出SWE-Shepherd框架，通过引入过程奖励模型为代码代理提供密集的步骤级监督，提升代码代理在大型代码库中的交互效率和动作质量。

Comments Code is available at https://github.com/mahirlabibdihan/swe-shepherd

HetRL: Efficient Reinforcement Learning for LLMs in Heterogeneous Environments

HetRL：异构环境下高效的大语言模型强化学习

Yongjun He, Shuai Zhang, Jiading Gai, Xiyuan Zhang, Boran Han, Bernie Wang, Huzefa Rangwala, George Karypis

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；post-training(abstract)

AI总结 本文提出HetRL系统，通过混合调度算法和整数线性规划方法，解决异构GPU环境下大语言模型强化学习的调度问题，实验表明其在吞吐量上优于现有系统。

Comments To appear at MLSys 2026

SYN-DIGITS: A Synthetic Control Framework for Calibrated Digital Twin Simulation

SYN-DIGITS: 一种合成控制框架用于校准数字孪生模拟

Grace Jiarui Fan, Chengpiao Huang, Tianyi Peng, Kaizheng Wang, Yuhang Wu

机构 * Finance Division, Columbia Business School（哥伦比亚商学院金融系） ； Department of IEOR, Columbia University（哥伦比亚大学工业工程与运筹学系） ； Decision, Risk, and Operations Division, Columbia Business School（哥伦比亚商学院决策、风险与运营系） ； Department of IEOR and Data Science Institute, Columbia University（哥伦比亚大学工业工程与运筹学系和数据科学研究所） 纠错

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

AI总结 本文提出SYN-DIGITS框架，通过学习数字孪生响应的潜在结构，校准大语言模型预测与人类行为的一致性，提升市场研究和社交科学中的模拟可靠性。

LASER: A Data-Centric Method for Low-Cost and Efficient SQL Rewriting based on SQL-GRPO

LASER: 一种数据驱动的低成本高效SQL重写方法基于SQL-GRPO

Jiahui Li, Tongwang Wu, Yuren Mao, Rong Kang, Tieying Zhang, Yunjun Gao

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；small language model(abstract)

AI总结 本文提出LASER，一种数据驱动的SQL重写方法，通过构建SQL-MCTS大规模语料库和SQL-GRPO专用对齐策略，提升小模型在SQL优化中的执行效率和零样本迁移能力。

Weakly Supervised Distillation of Hallucination Signals into Transformer Representations

弱监督蒸馏 hallucination 信号到 Transformer 表示

Shoaib Sadiq Salehmohamed, Jinal Prashant Thakkar, Hansika Aredla, Shaik Mohammed Omar, Shalmali Ayachit

机构 * LLM Lens 纠错

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

AI总结 本文提出弱监督框架，通过子串匹配、句子嵌入相似性和LLM判官信号，无需人工标注即可标注生成响应是否 grounded 或 hallucinated，并通过五种探针分类器验证 hallucination 信号可被蒸馏到 Transformer 表示中。

Comments 20 pages, 6 figures, 6 tables. Introduces a 15k-sample representation-level hallucination dataset with full transformer hidden states and multi-signal weak supervision. Evaluates 5 probing architectures and demonstrates internal hallucination detection without external inference-time signals. Includes held-out test evaluation and deployment benchmarks

LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification

LongSpec: 长上下文无损推测解码与高效草稿生成与验证

Penghui Yang, Cunxiao Du, Fengzhuo Zhang, Haonan Wang, Tianyu Pang, Chao Du, Bo An

机构 * Sea AI Lab（Sea AI实验室） ； Nanyang Technological University（南洋理工大学） ； National University of Singapore（新加坡国立大学） 纠错

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

AI总结 LongSpec通过三种创新解决长上下文场景下的推测解码问题，实现3.26倍速度提升和2.25倍时间减少，适用于长上下文理解任务。

Comments Accepted by ACL'25 (Main)

In-Place Test-Time Training

就地测试时间训练

Guhao Feng, Shengjie Luo, Kai Hua, Ge Zhang, Di He, Wenhao Huang, Tianle Cai

机构 * Peking University（北京大学） 纠错

专题命中 效率与部署 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL、cs.AI、cs.LG

AI总结 本文提出了一种无需重新训练的就地测试时间训练框架，通过调整MLP块的最终投影矩阵实现模型在推理时的动态更新，提升大语言模型在长上下文任务中的性能。

Comments ICLR 2026 Oral Presentation; Code is released at https://github.com/ByteDance-Seed/In-Place-TTT