01政策相关
102行业前沿动态
7Claude Platform 正式上线 Computer Use、Skills API 与 Files API，新增浏览器操作工具
22 天前Anthropic 宣布 Computer Use、Skills API 与 Files API 在 Claude Platform 全面可用，并新增浏览器操作工具，让智能体可操作软件、调用团队技能并返回成品文件。
Anthropic 如何开展 AI 教学
22 天前Anthropic 发布 Claude Academy，为全球数百万用户提供 AI 教学资源，帮助其安全、有效地使用 AI。该学院课程借鉴其内部员工培训方法，包括 4D AI Fluency Framework 及"ever-boarding"持续学习项目，并强调以问题为中心、培养持久思维模式而非特定操作行为。
Hugging Face 发布 LFM2.5 系列 DSpark 草稿模型，推理速度最高提升 3.18 倍
22 天前Hugging Face 发布 LFM2.5 系列三款模型的 DSpark 草稿模型检查点，通过投机解码在不改变输出质量的前提下，GPU 吞吐最高提升 3.18 倍，端侧最高 2.87 倍。草稿模型约 300M 参数，LFM2.5-2.6B 函数调用延迟平均降低 57%，已开源支持 llama.cpp 和 SGLang。
Mistral 推出 Agentic Search：多步检索提升 AI 系统复杂文档查询准确率
22 天前Mistral 发布 Agentic Search，通过 search、open、navigate、read、grep 五工具的多步检索循环，让模型在长文档与多来源中查找、定位并验证信息。
AlloyDB ScaNN 如何将向量搜索扩展到 100 亿向量
22 天前AlloyDB 的 ScaNN 索引现已支持超过 100 亿向量的规模，通过全新的四层树架构（预览版）实现，将查询复杂度从 O（N^1/2） 降至 O（N^1/4）。内部测试中，该架构在 100 亿向量规模下可实现 p95 延迟不超过 51 毫秒、召回率达 95%。该功能可通过快速入门指南部署，新用户可享受 30 天免费试用。
阿里发布 Qwen-UI-Agent，主打让模型真正"会用"每一块屏幕
22 天前阿里巴巴正式推出 Qwen-UI-Agent，一个以真实世界为中心的 GUI 智能体基座模型，覆盖移动端、电脑端、网页端及深度搜索（DeepSearch）环境。
Claude Platform 发布 Python SDK v1.0，迁移至 httpx2
23 天前Anthropic 发布 Claude Python SDK v1.0，HTTP 层从 httpx 迁移至 API 兼容的 httpx2，并移除 Text Completions API 等长期弃用功能。
04论文研究
8LEDGER: Claim-to-Evidence Trace Graphs for Auditing LLM Agents
22 天前arXiv:2608.18398v1 Announce Type: new Abstract: Large language model (LLM) agents can now carry out long-horizon technical workflows involving complex tool use, code execution, file edits, and generated artifacts. As age…
Multimodal Rapport Estimation in Real-World HRI
22 天前arXiv:2608.18401v1 Announce Type: new Abstract: Evaluating interaction quality in real-world HRI is an important challenge. If interaction quality can be estimated reliably, the results can be used to improve dialogue st…
Reducing Technician Search Burden: A Multimodal RAG for Cessna 172 Maintenance Manual
22 天前arXiv:2608.18465v1 Announce Type: new Abstract: Proper use of the aircraft maintenance manual is essential for correct maintenance, providing procedures, diagrams, cautions, and specifications. However, technicians often…
Designing Social Robots for Social-Cognition Training with Autistic Adults
22 天前arXiv:2608.18488v1 Announce Type: new Abstract: Social robots have been widely explored as tools for autism intervention, yet this literature has focused predominantly on children and has rarely involved autistic adults …
Measuring Proof Burden in Public Bounty Listings: A RentAHuman Case Study
22 天前arXiv:2608.18547v1 Announce Type: new Abstract: Online bounty markets let requesters advertise paid tasks. Workers may be asked not just to complete a task but to prove it, and proof can mean exposure: revealing identity…
SemanticSlider3D: Training-Free Continuous Semantic Editing for 3D Objects
22 天前arXiv:2608.18560v1 Announce Type: new Abstract: Fine-grained control over continuous semantic attributes of 3D objects is essential for 3D content creation, but is not well supported by conventional 3D modeling workflows…
Report on The 1st Workshop on Human-Centered Proactive and Personalized Agents for Interactive Information Access at CHI…
22 天前arXiv:2608.18638v1 Announce Type: new Abstract: Interactive information access is increasingly moving beyond reactive query-response paradigms toward agentic systems that can personalize interaction, retain context, infe…
Sounds Uncertain: Exploring the Affective Aspects of Sonification for Uncertainty Visualization
22 天前arXiv:2608.18680v1 Announce Type: new Abstract: Affective visualization can influence how users perceive, interpret, and engage with data by embedding and conveying emotion through visual design. While sound is widely us…