1
GeoRA: 为RLVR设计的LoRA——ACL 2026杰出论文解析
ACL 2026杰出论文奖（Outstanding Paper）揭晓，全球共 18 篇入选，其中包含美团履约技术团队的1篇论文。本文介绍了一种专为 RLVR 设计的低秩训练方法，以及它在业务 Agentic RL 中的落地经验。
2
Training Proactive and Personalized LLM Agents
让大模型从“被动应答”走向“主动服务”，教你训练会察言观色的个性化AI代理，附完整方法框架。
arXiv:2511.02208v2 Announce Type: replace Abstract: Despite rapid progress, current AI agents are primarily optimized for isolated task completion. We…
3
研究团队调查发现超七成英语生物医学论文已使用 AI 辅助写作，呼吁业界建立完善监管机制
精准识别生物医学论文中的AI辅助写作痕迹，为学术诚信提供数据支撑，多引擎检测更可靠
IT之家 8 月 25 日消息，据外媒 phys 报道，近期 Lena Holzwarth 带头的研究团队分析了来自 PubMed Central 平台的 119 万余篇英文生物医学论文，发现生成式 AI 正迅速融入学术论文写作。 研究人员通过分析论文中的“AI 高频用词”（例“delves”“ex…
4
CAI-DLLM: Convergence Aware Inference for Diffusion Language Models
扩散语言模型推理提速新思路，用收敛感知机制减少计算浪费，值得关注。
arXiv:2608.22646v1 Announce Type: new Abstract: Diffusion language models can generate many tokens in parallel, but they still require repeated denois…
5
6
Share the Judge, Learn the Deferral: Where Specialization Helps LLM Evaluation
揭秘专用裁判与共享评判的取舍之道，为LLM评估定制最优策略，实验详实、图表清晰。
arXiv:2607.27984v2 Announce Type: replace Abstract: Agentic systems generate outputs faster than human review. We contrast two LLM evaluator specializ…
7
Asymmetric Capacity Allocation in Self-Refinement Pipelines
探索自精炼流程中非对称容量分配策略，揭秘资源优化与推理效率的关键突破。
arXiv:2608.21345v1 Announce Type: new Abstract: Self-refinement, typically structured as generation, critique, and revision, is a widely adopted parad…
8
Mind the Style: Impact of Communication Style on Human-Chatbot Interaction
读这篇论文，搞懂聊天机器人沟通风格如何左右用户体验与任务成功率，交互设计必读
arXiv:2602.17850v2 Announce Type: replace-cross Abstract: Conversational agents increasingly mediate everyday digital interactions, yet the effects of…
9
Show HN: Circuit Breaker – Score Pull Requests
用论文算法给PR打分，提交前就能预判合并风险，省时省力。
Creation-time PR risk triage GitHub Action, based on Minh et al. MSR '26 (arXiv:2601.00753) Comments URL: https://news.ycombinator.com/item?id=4939113…
10
Break It Down, Pass It On: Cross-Task Skill Transfer in LLM Agents
从分解到传递，这项研究让LLM智能体把技能迁移到新任务，少训练、更高效，跨任务学习的进阶思路。
arXiv:2608.20274v1 Announce Type: cross Abstract: Large language model (LLM) agents can induce skills from completed tasks and reuse them later to gro…
11
Projector Is All You Train
投影器成训练核心？这篇论文挑战传统微调范式，或为高效训练提供新思路。
arXiv:2608.19726v1 Announce Type: cross Abstract: The typical training process of a multimodal large language model (MLLM) involves adapting both the …
12
The Lifecycle of LLM-as-a-Judge for Large-Scale Recommendation Explanations
揭秘LLM裁判在推荐解释中的全生命周期，从训练到部署的工程实践与挑战一网打尽。
arXiv:2608.18300v1 Announce Type: new Abstract: LLM-as-a-Judge, which leverages a large language model to evaluate natural language generated by anoth…
13
A Theory of Post-hoc Debate Judgement
从理论层面拆解事后辩论评判机制，给AI对齐与评估研究提供全新视角。
arXiv:2608.19002v1 Announce Type: new Abstract: Debates have recently emerged as a useful methodology for agentic AI to improve performance as well as…
14
Uncertainty-Aware Decision Making in Multimodal Large Language Models
多模态大模型如何量化不确定性？这项研究为高风险场景下的可靠决策提供新思路。
arXiv:2608.17084v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) increasingly answer questions whose correctness depends on vi…
15
When AI Designs AI: Innovation or Imitation?
AI自我进化的边界在哪？这篇论文直面“AI设计AI”是创新还是模仿的终极拷问。
arXiv:2608.17471v1 Announce Type: new Abstract: Recent advances in LLM agents have made them increasingly capable of designing methods for complex AI …
16
Looking for an ArXiv Endorser – AI / Computer Science
AI/CS研究者快来帮把手，作者正在为论文寻找ArXiv背书人，机会难得。
Article URL: https://github.com/paibyun9/EGA-V9 Comments URL: https://news.ycombinator.com/item?id=49357141 Points: 1 # Comments: 2
17
Semantic Uncertainty-Guided Orchestration in Hierarchical Multi-Agent Systems
从语义不确定性切入，为层级多智能体协作提供全新编排思路，适合关注大模型Agent与系统优化的读者。
arXiv:2608.14707v1 Announce Type: new Abstract: As large language model (LLM)-based multi-agent systems become increasingly capable, coordinating agen…
18
Pallas: A Proactive KV Cache Migration Framework for LLM Inference in AI-RAN
首个面向AI-RAN的主动式KV缓存迁移框架，直击LLM推理中长序列与多边缘节点的时延痛点，架构设计值得关注。
arXiv:2608.16477v1 Announce Type: new Abstract: AI-RAN brings large language model (LLM) serving close to mobile users, but cellular handover can sepa…
19
Understanding Cognition-Induced Risks in Agentic AI Systems
智能体AI系统因“认知”而产生哪些独特风险？这篇论文给出系统化剖析框架。
arXiv:2608.15304v1 Announce Type: new Abstract: Frontier agentic systems powered by large language models (LLMs) exhibit human-like patterns of cognit…
20
Towards Risk-free AI Agent Deployment
AI代理落地如何规避风险？这篇论文提出系统化部署框架，直击安全与可靠性痛点。
arXiv:2608.16411v1 Announce Type: cross Abstract: LLM-based agents are rapidly moving from research prototypes into the core business processes of org…