语言大模型 / LLM

Towards Continuous Intelligence Growth: Self-Training, Continual Learning, and Dual-Scale Memory in SuperIntelliAgent

迈向持续智能增长：在SuperIntelliAgent中实现自我训练、持续学习和双尺度记忆

Jianzhe Lin, Zeyu Pan, Yun Zhu, Ruiqi Song, Jining Yang

专题命中 长上下文与记忆 ：large language model(abstract)；language model(abstract)；preference optimization(abstract)；分类 cs.AI

AI总结 SuperIntelliAgent通过自我训练和双尺度记忆实现持续智能增长，利用可训练学习者与推理验证者配对，提升偏好对齐和学习效果。

Comments 15 pages, 4 figures

More Documents, Same Length: Isolating the Challenge of Multiple Documents in RAG

更多文档，相同长度：隔离多文档在RAG中的挑战

Shahar Levy, Nir Mazor, Lihi Shalmon, Michael Hassid, Gabriel Stanovsky

机构 * School of Computer Science and Engineering（计算机科学与工程学院） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

AI总结 研究发现RAG中增加文档数量显著降低LLM性能，但Qwen2.5表现出更强的多文档处理能力。

Comments Preprint

AutoLink: Autonomous Schema Exploration and Expansion for Scalable Schema Linking in Text-to-SQL at Scale

AutoLink：用于大规模文本到SQL的自主模式探索与扩展

Ziyang Wang, Yuanlei Zheng, Zhenbiao Cao, Xiaojin Zhang, Zhongyu Wei, Pei Fu, Zhenbo Luo, Wei Chen, Xiang Bai

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

AI总结 AutoLink通过自主代理框架实现大规模文本到SQL的模式探索与扩展，提升模式链接的召回率和可扩展性。

Whistledown: Combining User-Level Privacy with Conversational Coherence in LLMs

Chelsea McMurray, Hayder Tirmazi

机构 * Dorcha（多查） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Fixed-Persona SLMs with Modular Memory: Scalable NPC Dialogue on Consumer Hardware

Martin Braas, Lukas Esterle

机构 * Andreasen 1（安德森实验室） ； Lukas Esterle 1,2（埃斯特勒实验室） 纠错

专题命中 长上下文与记忆 ：large language model(abstract)；language model(abstract)；small language model(abstract)；分类 cs.AI

Bridging LMS and generative AI: dynamic course content integration (DCCI) for enhancing student satisfaction and engagement via the ask ME assistant

Kovan Mzwri, Márta Turcsányi-Szabo

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Journal ref J. Comput. Educ. (2025)

Controllable Hybrid Captioner for Improved Long-form Video Understanding

Kuleen Sasse, Efsun Sarioglu Kayi, Arun Reddy

机构 * Johns Hopkins University Applied Physics Laboratory（约翰霍普金斯大学应用物理实验室） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

The Narrative Continuity Test: A Conceptual Framework for Evaluating Identity Persistence in AI Systems

Stefano Natangelo

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 33 pages, 127 references v2: Minor editorial revision: redundant phrasing reduced, punctuation and formatting improved; no conceptual or data changes

SABlock: Semantic-Aware KV Cache Eviction with Adaptive Compression Block Size

Jinhan Chen, Jianchun Liu, Hongli Xu, Xianjun Gao, Shilong Wang

机构 * School of Computer Science and Technology, University of Science and Technology of China（计算机科学与技术学院，科学技术大学） ； Suzhou Institute for Advanced Research, University of Science and Technology of China（苏州先进研究院，科学技术大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Modeling realistic human behavior using generative agents in a multimodal transport system: Software architecture and Application to Toulouse

Trung-Dung Vu, Benoit Gaudou, Kamaldeep Singh Oberoi

机构 * UMR IRIT University Toulouse Capitole（IRIT大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

The Right to Be Remembered: Preserving Maximally Truthful Digital Memory in the Age of AI

Alex Zhavoronkov, Dominika Wilczok, Roman Yampolskiy

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

STARK: Strategic Team of Agents for Refining Kernels

Juncheng Dong, Yang Yang, Tao Liu, Yang Wang, Feng Qi, Vahid Tarokh, Kaushik Rangadurai, Shuang Yang

机构 * Meta Ranking AI Research（Meta 排名人工智能研究） ； Duke University（杜克大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Leveraging Coordinate Momentum in SignSGD and Muon: Memory-Optimized Zero-Order

Egor Petrov, Grigoriy Evseev, Aleksey Antonov, Andrey Veprikov, Nikolay Bushkov, Stanislav Moiseev, Aleksandr Beznosikov

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments 26 pages, 5 tables

MemGen: Weaving Generative Latent Memory for Self-Evolving Agents

Guibin Zhang, Muxin Fu, Shuicheng Yan

机构 * National University of Singapore（新加坡国立大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Judgment of Learning: A Human Ability Beyond Generative Artificial Intelligence

Markus Huff, Elanur Ulakçı

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 24 pages, 2 figures

Journal ref Scientific Reports, 15(1), 35030 (2025)

Lost in the Middle: An Emergent Property from Information Retrieval Demands in LLMs

Nikolaus Salvatore, Hao Wang, Qiong Zhang

机构 * Rutgers University–New Brunswick（罗格斯大学新布朗斯维爾分校） ； Rutgers Center for Cognitive Science（罗格斯认知科学中心） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

ROBOPSY PL[AI]: Using Role-Play to Investigate how LLMs Present Collective Memory

Margarete Jahrmann, Thomas Brandstetter, Stefan Glasauer

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 17 pages, 4 figures

Implementing Semantic Join Operators Efficiently

Immanuel Trummer

机构 * Cornell University（康奈尔大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Mem-α: Learning Memory Construction via Reinforcement Learning

Yu Wang, Ryuichi Takanobu, Zhiqi Liang, Yuzhen Mao, Yuanzhe Hu, Julian McAuley, Xiaojian Wu

机构 * Anuttacon ； University of California San Diego（加州大学圣地亚哥分校） ； Stanford University（斯坦福大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

KVzip: Query-Agnostic KV Cache Compression with Context Reconstruction

Jang-Hyun Kim, Jinuk Kim, Sangwoo Kwon, Jae W. Lee, Sangdoo Yun, Hyun Oh Song

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments NeurIPS 2025 Oral. Code: https://github.com/snu-mllab/KVzip

From Single to Multi-Granularity: Toward Long-Term Memory Association and Selection of Conversational Agents

Derong Xu, Yi Wen, Pengyue Jia, Yingyi Zhang, wenlin zhang, Yichao Wang, Huifeng Guo, Ruiming Tang, Xiangyu Zhao, Enhong Chen, Tong Xu

机构 * University of Science and Technology of China（中国科学技术大学） ； City University of Hong Kong（香港城市大学） ； Huawei Noah’s Ark Lab（华为诺亚实验室） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

EvoMail: Self-Evolving Cognitive Agents for Adaptive Spam and Phishing Email Defense

Wei Huang, De-Tian Chu, Lin-Yuan Bai, Wei Kang, Hai-Tao Zhang, Bo Li, Zhi-Mo Han, Jing Ge, Hai-Feng Lin

机构 * People’s Liberation Army 77606 Unit（中国人民解放军77606单位） ； Field Engineering College, Army Engineering University of PLA（陆军工程大学工程学院） ； College Of Software, ZhengZhou University of Light Industry（郑州轻工业大学软件学院） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Meta-Memory: Retrieving and Integrating Semantic-Spatial Memories for Robot Spatial Reasoning

Yufan Mao, Hanjing Ye, Wenlong Dong, Chengjie Zhang, Hong Zhang

机构 * Southern University of Science and Technology（南方科技大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Benchmarking ChatGPT and DeepSeek in April 2025: A Novel Dual Perspective Sentiment Analysis Using Lexicon-Based and Deep Learning Approaches

Maryam Mahdi Alhusseini, Mohammad-Reza Feizi-Derakhshi

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 17 pages, 21 figures

REMI: A Novel Causal Schema Memory Architecture for Personalized Lifestyle Recommendation Agents

Vishal Raman, Vijai Aravindh R, Abhijith Ragav

机构 * Radian Group Inc.（Radian集团） ； Sri Sivasubramaniya Nadar College Of Engineering（Sri Sivasubramaniya纳达尔工程学院） ； Amazon（亚马逊） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 8 pages, 2 figures, Accepted at the OARS Workshop, KDD 2025, Paper link: https://oars-workshop.github.io/papers/Raman2025.pdf

Modeling shopper interest broadness with entropy-driven dialogue policy in the context of arbitrarily large product catalogs

Firas Jarboui, Issa Memari

机构 * Gorgias Paris France（巴黎高斯集团） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Using LLMs to create analytical datasets: A case study of reconstructing the historical memory of Colombia

David Anderson, Galia Benitez, Margret Bjarnadottir, Shriyan Reyya

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Joint Enhancement of Relational Reasoning for Long-Context LLMs

Zhirui Chen, Wei Shen, Jiashui Huang, Ling Shao

机构 * UCAS-Terminus AI Lab, University of Chinese Academy of Sciences, China（中国科学院大学Terminus AI实验室，中国科学院大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 9 pages, 5 pages Accepted by EMNLP 2025 Findings

SwizzlePerf: Hardware-Aware LLMs for GPU Kernel Performance Optimization

Arya Tschand, Muhammad Awad, Ryan Swann, Kesavan Ramakrishnan, Jeffrey Ma, Keith Lowery, Ganesh Dasika, Vijay Janapa Reddi

机构 * Harvard University（哈佛大学） ； AMD ； Stanford University（斯坦福大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Insights into User Interface Innovations from a Design Thinking Workshop at deRSE25

Maximilian Frank, Simon Lund

机构 * Ludwig-Maximilians-Universität München（慕尼黑路易斯-马克西米利安大学） 纠错

专题命中 长上下文与记忆 ：LLM(abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI