语言大模型 / LLM

AdaptFuse: Training-Free Sequential Preference Learning via Externalized Bayesian Inference

AdaptFuse：通过外部化贝叶斯推断实现无需训练的序列偏好学习

Fangzhou Lin, Peiran Li, Shuo Xing, Siyuan Yang, Qianwen Ge, Kazunori Yamada, Ziming Zhang, Haichong Zhang, Zhengzhong Tu

机构 * Texas A&M University（德克萨斯农工大学） ； Worcester Polytechnic Institute（伍斯特理工学院） ； Tohoku University（东北大学） ； Georgia Institute of Technology（佐治亚理工学院） 纠错

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

AI总结 AdaptFuse通过外部化贝叶斯推断实现无需训练的序列偏好学习，利用符号模块和冻结LLM结合熵自适应融合，提升推荐系统性能，无需敏感用户数据训练。

Comments 20 pages, 4 figures, 5 tables

Controllable and explainable personality sliders for LLMs at inference time

用于推理时的可控且可解释的人格滑块

Florian Hoppe, David Khachaturov, Robert Mullins, Mark Huasong Meng

机构 * Technical University of Munich, Germany（慕尼黑技术大学） ； University of Cambridge, United Kingdom（剑桥大学） ； University College Dublin, Ireland（都柏林大学） 纠错

专题命中 效率与部署 ：large language model(abstract)；language model(abstract)；SFT(abstract)；RLHF(abstract)

AI总结 本文提出一种模块化框架，通过顺序自适应引导方法实现LLM推理时的可控且可解释的人格控制，通过正交化引导向量提升多维人格调节的精度和效率。

Comments 20 pages, 18 figures

Inference-Time Safety For Code LLMs Via Retrieval-Augmented Revision

通过检索增强的修订实现代码LLM的推理时安全性

Manisha Mukherjee, Vincent J. Hellendoorn

机构 * Carnegie Mellon University（卡内基梅隆大学） 纠错

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

AI总结 通过检索增强的修订机制提升代码LLM的推理时安全性，提高生成代码的安全性并减少漏洞。

Comments Accepted at the ICLR 2026 Workshop on Principled Design for Trustworthy AI: Interpretability, Robustness, and Safety Across Modalities

Goal Inference from Open-Ended Dialog

从开放式对话中推断目标

Rachel Ma, Jingyi Qu, Andreea Bobu, Dylan Hadfield-Menell

机构 * MIT CSAIL（麻省理工学院计算机科学与人工智能实验室） 纠错

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

AI总结 本研究提出了一种在线方法，通过自然语言提取目标并利用贝叶斯推断，使具身代理在开放式对话中高效学习和完成多样化用户目标。

Comments This version has been updated to reflect a copy of Master's thesis submitted Jan 24, 2025 for degree date Feb 2025 (https://hdl.handle.net/1721.1/158960). We recommend readers to read revised version incorporating a different agent pipeline and methodological approach which is available at: arXiv:2508.15119

Train Less, Learn More: Adaptive Efficient Rollout Optimization for Group-Based Reinforcement Learning

少训练，多学习：基于群体的强化学习中的自适应高效回滚优化

Zhi Zhang, Zhen Han, Costas Mavromatis, Qi Zhu, Yunyi Zhang, Sheng Guan, Dingmin Wang, Xiong Zhou, Shuai Wang, Soji Adeshina, Vassilis Ioannidis, Huzefa Rangwala

机构 * University of California, Los Angeles（加州大学洛杉矶分校） ； Amazon Web Services（亚马逊网络服务） 纠错

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

AI总结 AERO通过自适应回滚策略和选择性拒绝提升群体强化学习的计算效率，减少48%的训练计算并缩短45%的训练时间，同时保持或提升性能。

On the Importance of a Multi-Scale Calibration for Quantization

关于多尺度校准在量化中的重要性

Seungwoo Son, Ingyu Seong, Junhan Kim, Hyemi Jang, Yongkweon Jeon

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MaCa方法，通过多尺度校准提升大语言模型量化精度，改进Hessian估计并增强量化效果。

Comments ICASSP 2026

Semantic Search At LinkedIn

LinkedIn 的语义搜索

Fedor Borisyuk, Sriram Vasudevan, Muchen Wu, Guoyao Li, Benjamin Le, Shaobo Zhang, Qianqi Kay Shen, Yuchin Juan, Kayhan Behdin, Liming Dong, Kaixu Yang, Shusen Jing, Ravi Pothamsetty, Rajat Arora, Sophie Yanying Sheng, Vitaly Abdrashitov, Yang Zhao, Lin Su, Xiaoqing Wang, Chujie Zheng, Sarang Metkar, Rupesh Gupta, Igor Lapchuk, David N. Racca, Madhumitha Mohan, Yanbo Li, Haojun Li, Saloni Gandhi, Xueying Lu, Chetan Bhole, Ali Hooshmand, Xin Yang, Raghavan Muthuregunathan, Jiajun Zhang, Mathew Teoh, Adam Coler, Abhinav Gupta, Xiaojing Ma, Sundara Raman Ramachandran, Morteza Ramezani, Yubo Wang, Lijuan Zhang, Richard Li, Jian Sheng, Chanh Nguyen, Yen-Chi Chen, Chuanrui Zhu, Claire Zhang, Jiahao Xu, Deepti Kulkarni, Qing Lan, Arvind Subramaniam, Ata Fatahibaarzi, Steven Shimizu, Yanning Chen, Zhipeng Wang, Ran He, Zhengze Zhou, Qingquan Song, Yun Dai, Caleb Johnson, Ping Liu, Shaghayegh Gharghabi, Gokulraj Mohanasundaram, Juan Bottaro, Santhosh Sachindran, Qi Guo, Yunxiang Ren, Chengming Jiang, Di Mo, Luke Simon, Jianqiang Shen, Jingwei Wu, Wenjing Zhang

机构 * LinkedIn（领英） 纠错

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 LinkedIn推出基于大语言模型的语义搜索框架，通过优化相关性和参与度，在保持效率的同时提升检索质量和用户参与度。

FAQ: Mitigating Quantization Error via Regenerating Calibration Data with Family-Aware Quantization

FAQ: 通过家庭感知量化减少量化误差

Haiyang Xiao, Weiqing Li, Jinyue Guo, Guochao Jiang, Guohua Liu, Yuewei Zhang

机构 * Alibaba Cloud Computing（阿里云计算） 纠错

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

AI总结 FAQ通过家庭感知量化方法生成高保真校准数据，有效减少量化误差，实验表明其在多个模型系列中将准确性损失降低28.5%。

AutoToM: Scaling Model-based Mental Inference via Automated Agent Modeling

AutoToM: 通过自动化代理建模实现基于模型的心理推理扩展

Zhining Zhang, Chuanyang Jin, Mung Yao Jia, Shunchi Zhang, Tianmin Shu

机构 * Peking University（北京大学） ； Johns Hopkins University（约翰霍普金斯大学） 纠错

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

AI总结 AutoToM通过自动化代理建模实现可扩展、稳健且可解释的心理推理，优于现有方法并支持在线心理推理。

Comments NeurIPS 2025 (Spotlight). 42 pages, 11 figures, 15 tables. Website at https://chuanyangjin.com/AutoToM/

Phishing Detection in the Gen-AI Era: Quantized LLMs vs Classical Models

在生成式AI时代进行钓鱼攻击检测：量化LLM与经典模型

Jikesh Thapa, Gurrehmat Chahal, Serban Voinea Gabreanu, Yazan Otoum

机构 * School of Computer Science and Technology（计算机科学与技术学院） 纠错

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

AI总结 本文比较了传统ML、DL和量化LLM在钓鱼检测中的性能，发现优化的LLM在准确率和效率上具有潜力，适合高效部署。

Comments 8 Pages, IEEE Conference

Efficient Text Classification with Conformal In-Context Learning

高效文本分类中的符合性上下文学习

Ippokratis Pantelidis, Korbinian Randl, Aron Henriksson

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CICLe框架，通过结合轻量级分类器与符合性预测，提升文本分类效率，减少样本和提示长度，尤其在高类别不平衡任务中表现优异。

Comments 10 pages, 4 tables, 2 figures

From Pixels to Posts: Retrieval-Augmented Fashion Captioning and Hashtag Generation

从像素到帖子：基于检索的时尚描述与标签生成

Moazzam Umer Gondal, Hamad Ul Qudous, Daniya Siddiqui, Asma Ahmad Farhan

机构 * organization= School of Computing, National University of Computer \& Emerging Sciences (FAST) , city= Lahore , postcode= 54000 , country= Pakistan 纠错

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

AI总结 本文提出一种基于检索的时尚描述与标签生成方法，结合多件服装检测、属性推理和大语言模型，提升描述准确性和标签生成的视觉相关性。

Comments Submitted to Expert Systems with Applications

Bi-Mamba: Towards Accurate 1-Bit State Space Models

Shengkun Tang, Liqun Ma, Haonan Li, Mingjie Sun, Zhiqiang Shen

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

Comments Accepted in TMLR 2025

Unbiased Gradient Low-Rank Projection

Rui Pan, Yang Luo, Yuxing Liu, Yang You, Tong Zhang

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； 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)；pretraining(abstract)

Improving Factuality in LLMs via Inference-Time Knowledge Graph Construction

Shanglin Wu, Lihui Liu, Jinho D. Choi, Kai Shu

机构 * Emory University（埃默里大学） ； Wayne State University（韦恩州立大学） 纠错

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

SATER: A Self-Aware and Token-Efficient Approach to Routing and Cascading

Yuanzhe Shen, Yide Liu, Zisu Huang, Ruicheng Yin, Xiaoqing Zheng, Xuanjing Huang

机构 * School of Computer Science, Fudan University（复旦大学计算机学院） 纠错

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

Comments Accepted to EMNLP 2025 Main

Optimal Stopping vs Best-of-$N$ for Inference Time Optimization

Yusuf Kalayci, Vinod Raman, Shaddin Dughmi

机构 * University of Southern California（南加州大学） ； University of Michigan（密歇根大学） 纠错

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

Comments 24 pages

End-to-End On-Device Quantization-Aware Training for LLMs at Inference Cost

Qitao Tan, Xiaoying Song, Jin Lu, Guoming Li, Jun Liu, Lingzi Hong, Caiwen Ding, Jundong Li, Xiaoming Zhai, Shaoyi Huang, Wei Niu, Geng Yuan

机构 * University of Georgia（佐治亚大学） ； University of North Texas（北卡罗来纳州立大学） ； Northeastern University（东北大学） ； University of Minnesota（明尼苏达大学） ； University of Virginia（弗吉尼亚大学） ； Stevens Institute of Technology（史蒂文斯理工学院） 纠错

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

MiniCPM4: Ultra-Efficient LLMs on End Devices

MiniCPM Team, Chaojun Xiao, Yuxuan Li, Xu Han, Yuzhuo Bai, Jie Cai, Haotian Chen, Wentong Chen, Xin Cong, Ganqu Cui, Ning Ding, Shengda Fan, Yewei Fang, Zixuan Fu, Wenyu Guan, Yitong Guan, Junshao Guo, Yufeng Han, Bingxiang He, Yuxiang Huang, Baoxi Ji, Cunliang Kong, Qiuzuo Li, Siyuan Li, Wenhao Li, Xin Li, Yanghao Li, Yishan Li, Zhen Li, Dan Liu, Biyuan Lin, Yankai Lin, Xiang Long, Quanyu Lu, Yaxi Lu, Peiyan Luo, Hongya Lyu, Litu Ou, Yinxu Pan, Lushi Pu, Zekai Qu, Qundong Shi, Zijun Song, Jiayuan Su, Zhou Su, Ao Sun, Xianghui Sun, Peijun Tang, Fangzheng Wang, Feng Wang, Shuo Wang, Yudong Wang, Zheng Wang, Yesai Wu, Zhenyu Xiao, Jie Xie, Zihao Xie, Xiaoyue Xu, Yukun Yan, Jiarui Yuan, Jinqian Zhang, Kaihuo Zhang, Lei Zhang, Linyue Zhang, Xueren Zhang, Yudi Zhang, Hengyu Zhao, Weilin Zhao, Weilun Zhao, Yuanqian Zhao, Zhi Zheng, Chuyue Zhou, Ge Zhou, Jie Zhou, Wei Zhou, Yanghao Zhou, Zihan Zhou, Zixuan Zhou, Zhiyuan Liu, Guoyang Zeng, Chao Jia, Dahai Li, Maosong Sun

机构 * MiniCPM Team（MiniCPM团队） 纠错

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

Comments MiniCPM4 Technical Report

HypER: Literature-grounded Hypothesis Generation and Distillation with Provenance

Rosni Vasu, Chandrayee Basu, Bhavana Dalvi Mishra, Cristina Sarasua, Peter Clark, Abraham Bernstein

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

Comments EMNLP 2025, 26 pages (9 pages: main paper body)

Can Large Models Teach Student Models to Solve Mathematical Problems Like Human Beings? A Reasoning Distillation Method via Multi-LoRA Interaction

Xinhe Li, Jiajun Liu, Peng Wang

机构 * School of Computer Science and Engineering, Southeast University（东南大学计算机科学与工程学院） ； Key Laboratory of New Generation Artificial Intelligence Technology and Its Interdisciplinary Applications (Southeast University), Ministry of Education（新一代人工智能技术及其交叉应用重点实验室（东南大学），教育部） 纠错

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

Comments Accepted by IJCAI2025

REFN: A Reinforcement-Learning-From-Network Framework against 1-day/n-day Exploitations

Tianlong Yu, Lihong Liu, Ziyi Zhou, Fudu Xing, Kailong Wang, Yang Yang

机构 * School of Artificial Intelligence, Hubei University（湖北大学人工智能学院） ； Huazhong University of Science and Technology（华中科技大学） ； University of Southern California（美国南加州大学） 纠错

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

Motif 2.6B Technical Report

Junghwan Lim, Sungmin Lee, Dongseok Kim, Eunhwan Park, Hyunbyung Park, Junhyeok Lee, Wai Ting Cheung, Dahye Choi, Jaeheui Her, Jaeyeon Huh, Hanbin Jung, Changjin Kang, Beomgyu Kim, Jihwan Kim, Minjae Kim, Taehwan Kim, Youngrok Kim, Haesol Lee, Jeesoo Lee, Kungyu Lee, Dongpin Oh, Yeongjae Park, Bokki Ryu, Daewon Suh, Dongjoo Weon

机构 * Motif Technologies 纠错

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

Efficient Knowledge Injection in LLMs via Self-Distillation

Kalle Kujanpää, Pekka Marttinen, Harri Valpola, Alexander Ilin

机构 * Aalto University（阿alto大学） ； Finnish Center for Artificial Intelligence (FCAI)（芬兰人工智能中心） ； System 2 AI 纠错

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

Comments Preprint

NestQuant: Nested Lattice Quantization for Matrix Products and LLMs

Semyon Savkin, Eitan Porat, Or Ordentlich, Yury Polyanskiy

机构 * MIT（麻省理工学院） ； Hebrew University of Jerusalem（耶路撒冷希伯来大学） 纠错

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

Comments 23 pages; Accepted at the 42nd International Conference on Machine Learning (ICML 2025)

DailyLLM: Context-Aware Activity Log Generation Using Multi-Modal Sensors and LLMs

Ye Tian, Xiaoyuan Ren, Zihao Wang, Onat Gungor, Xiaofan Yu, Tajana Rosing

机构 * University of California San Diego, Computer Science and Engineering Department（加州大学圣地亚哥分校计算机科学与工程系） 纠错

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

Unifying Block-wise PTQ and Distillation-based QAT for Progressive Quantization toward 2-bit Instruction-Tuned LLMs

Jung Hyun Lee, Seungjae Shin, Vinnam Kim, Jaeseong You, An Chen

机构 * Qualcomm AI Research（高通人工智能研究） 纠错

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

Comments Preprint

Efficient Data Selection at Scale via Influence Distillation

Mahdi Nikdan, Vincent Cohen-Addad, Dan Alistarh, Vahab Mirrokni

机构 * ISTA ； Google Research（谷歌研究） ； Red Hat AI（红帽人工智能） 纠错

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

LARGO: Latent Adversarial Reflection through Gradient Optimization for Jailbreaking LLMs

Ran Li, Hao Wang, Chengzhi Mao

机构 * Columbia University（哥伦比亚大学） ； Rutgers University（罗格斯大学） 纠错

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

Compensate Quantization Errors+: Quantized Models Are Inquisitive Learners

Yifei Gao, Jie Ou, Lei Wang, Jun Cheng, Mengchu Zhou

机构 * Yifei Gao ∗ , Jie Ou ∗ , Lei Wang † † {\dagger} † , Jun Cheng, and Mengchu Zhou ∗（作者） 纠错

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

Comments Effecient Quantization Methods for LLMs