1
Tracing Agentic Failure from the Flow of Success
从成功流程中追溯智能体失败根源，揭示AI自主决策的脆弱性
arXiv:2607.12747v1 Announce Type: cross Abstract: Failure attribution for LLM-based agentic systems, i.e., identifying which steps in a failure trajec…
2
Index SLM Technical Report
arXiv 让你免费获取最新学术论文预印本，无需订阅即可一键下载 PDF，追踪前沿研究。
arXiv:2607.09885v1 Announce Type: new Abstract: We present Index-1.9B, a series of open small language models developed at Bilibili. The series compri…
3
VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI Agents
用AI代理自动发现并利用IoT漏洞，前沿安全研究论文。
arXiv:2607.09653v1 Announce Type: cross Abstract: Internet of Things (IoT) systems are inherently vulnerable due to constrained hardware, outdated fir…
4
VectorizationLLM: Smart Vectorization Based AI Assistant
基于谷歌开源LLM打造的专用向量化AI助手，论文详解模型设计与应用场景。
arXiv:2607.07846v1 Announce Type: new Abstract: VectorizationLLM is a specialized Large Language Model based on Google open-weight LLMs. The model is …
5
Beyond Reactivity: Measuring Proactive Problem Solving in LLM Agents
突破传统反应式评估，为LLM智能体主动问题解决能力提出全新测量框架。
arXiv:2510.19771v4 Announce Type: replace Abstract: LLM-based agents are increasingly moving towards proactivity: rather than awaiting instruction, th…
6
Depression Symptoms and Relational Patterns in 187k ChatGPT Histories
187万条ChatGPT对话揭示抑郁症状与关系模式，AI数据助力心理健康研究新突破。
arXiv:2607.05685v1 Announce Type: cross Abstract: Large language models are increasingly used as private, always-available conversational systems, but…
7
HAL: Inducing Human-likeness in LLMs with Alignment
新框架HAL通过对齐策略引导大模型展现更像人类的特质，视角独特。
arXiv:2601.02813v3 Announce Type: replace Abstract: Aligning language models to qualitative behavioral traits, such as human-likeness, remains difficu…
8
Hey, That's My Model! Introducing Chain & Hash, An LLM Fingerprinting Technique
用Chain & Hash给LLM打上独特指纹，防止他人盗用你的模型。
arXiv:2407.10887v4 Announce Type: replace-cross Abstract: Growing concerns over the theft and misuse of Large Language Models (LLMs) underscore the ne…
9
Diversity is the Strength of the AI Crowd
揭秘AI群体智慧的核心：多样性如何成为提升AI性能的关键力量，来自ICML 2026工作坊的前沿研究。
arXiv:2606.29661v1 Announce Type: new Abstract: Top AI forecasting systems are approaching superforecaster-level accuracy on future world events, but …
10
Can LLMs Imagine Moral Alternatives Beyond Binary Dilemmas?
探索大语言模型能否超越简单二元对立，生成更具弹性的道德替代方案。
arXiv:2606.31213v1 Announce Type: cross Abstract: As large language models (LLMs) are increasingly deployed as moral advisors and agents, they need to…
11
Categorizing Mathematical Concepts with LLM Voting Ensembles in Mathswitch
用LLM投票集成方法为数学概念自动分类，这篇已被CICM录用的论文或能突破传统分类瓶颈。
arXiv:2606.28815v1 Announce Type: cross Abstract: Mathswitch is an open-source project that imports mathematical concept records from sources such as …
12
ICML 2026 | 美团技术团队学术论文精选
ICML是机器学习领域最具影响力的国际顶级学术会议之一。大会旨在探讨机器学习未来发展所面临的关键挑战与核心问题，并通过征集和评估具有重要理论价值和实际影响的前沿研究成果，推动领域发展并引领未来研究方向。
13
Seven Security Challenges That Must be Solved in Cross-domain Multi-agent LLM Systems
从攻击面到防护框架，系统梳理跨域多智能体LLM系统必须面对的七大安全挑战。
arXiv:2505.23847v4 Announce Type: replace-cross Abstract: Large language models (LLMs) are rapidly evolving into autonomous agents that cooperate acro…
14
LiveClawBench: Benchmarking LLM Agents on Complex, Real-World Assistant Tasks
全新基准测试平台，专为评估LLM智能体在复杂现实助手任务中的表现而设计，填补现有评估空白。
arXiv:2604.13072v2 Announce Type: replace-cross Abstract: OpenClaw-style personal assistants extend LLM agents from isolated tool use to open-ended, s…
15
AI Healthcare Chatbots as Information Infrastructure: A Large-Scale Study of User-Reported Breakdowns
海量预印本论文免费获取，每日更新AI、医学等前沿研究，是快速追踪最新学术动态的必备资源库。
arXiv:2606.27302v1 Announce Type: cross Abstract: AI healthcare chatbots are increasingly used to support health information seeking and self-manageme…
16
Thinking Like a Scientist? A Structural Study of LLM-Generated Research Methods
评估大模型能否像科学家一样思考：从结构层面剖析LLM生成的研究方法，揭秘AI科研的成色与局限。
arXiv:2606.26130v1 Announce Type: cross Abstract: Large Language Models (LLMs) are increasingly used to guide research methodology, yet their default …
17
The Open Source Economic Index of AI Adoption and Capability
首个量化开源AI经济影响的学术指标，揭示全球AI采用与能力的地图
arXiv:2606.26118v1 Announce Type: cross Abstract: We work towards measuring both AI adoption and the capability of AI to perform discrete labor tasks …
18
LLM-based Models for Detecting Emerging Topics in Service Feedback
用LLM从海量服务反馈中自动发现新兴话题，为产品优化提供数据驱动的洞察
arXiv:2606.26595v1 Announce Type: new Abstract: Enhancing the analysis of service feedback is essential for public sector organizations, particularly …
19
Agentic Analysis for Agentic Infrastructure: An LLM-Powered Pipeline for Comparative Governance of DAO and Corporate AI Protocols
用LLM驱动管道分析DAO与公司AI协议的治理差异，代理基础设施新视角
arXiv:2606.26203v1 Announce Type: new Abstract: As AI agent protocols proliferate, the governance structures shaping their interoperability standards …
20
Refusal Lives Downstream of Persona in Chat Models
arXiv.org全球学术论文预印本平台，免费获取最新研究成果，开放共享加速科学传播。
arXiv:2606.26161v1 Announce Type: new Abstract: Linear directions in activation space have been identified for both refusal and persona traits in inst…