论文精选 82
WearableQA: A Benchmark for Health Reasoning over Real-World Wearable Data· 可穿戴健康推理基准
Recent advances in wearable sensing enable continuous monitoring of physiological and beha…
领域：cs.CL作者：Ji Soo Lee、Xilun Chen、Pierce Chuang
📰 AI 研究简报 · 9月8日
今日头条：GPT-6 Astra Changes Everything —— YouTube · Two Minute Papers
共 51 条预计阅读 31 分钟Token 消耗 91,800
arXiv 最新论文
论文精选 60已解读
A Deep Generative Model for Synthesizing Labeled Wireless Signals· 用于合成带标签无线信号的深度生成模型
Wireless signals with position-related labels are pivotal for both performance evaluation …
领域：cs.AI作者：Yuxiao Li、Keke Hu、Santiago Mazuelas
论文精选 60已解读
Multi-Step Tool-Calling over Korean Open Public APIs: A Benchmark and a Data-Synthesis Recipe· 韩国开放公共API的多步调用基准
Data-sovereignty regulations increasingly require public institutions to deploy open-sourc…
领域：cs.AI作者：Dain Kim、Eungi Cho、Kyumin Kim
论文精选 60已解读
Necessary or Sufficient? Evaluating LLM Explanations With Behavioural Evidence· 必要还是充分？用行为证据评估LLM解释
LLM decision components that can operate within agent workflows often produce action-relev…
领域：cs.AI作者：Urja Pawar、Rajitha Ramanayake、Nabeel Kemal
论文精选 60已解读
Molecular Déjà Vu: Digit-Level Retrieval of Published Values in Frontier Language Models· 分子属性预测中的大模型文字级检索
Large language models (LLMs) are increasingly evaluated on molecular property benchmarks, …
领域：cs.AI作者：Matthias Busch、Marius Tacke、Sviatlana V. Lamaka
论文精选 60已解读
CUA-Universe: A Scalable and Dynamic Environment for Hybrid GUI+CLI Agents· CUA-Universe: GUI+CLI 混合代理的可扩展动态环境
Computer-use agents have advanced on benchmarks like OSWorld and AndroidWorld, but still a…
领域：cs.AI作者：Haoting Shi、Wenhao Wang、Weicheng Fang
论文精选 60已解读
Who Should Grade My Work? Student Perspectives on Transparent AI-Assisted Writing Assessment in Higher Education· 谁来评价我的作业？学生对透明AI写作评估的观点
The integration of GenAI tools into higher education assessment raises important questions…
领域：cs.AI作者：Rayed AlGhamdi
论文精选 60已解读
Does Your Agent's Memory Survive a Model Upgrade? A Controlled Study of Memory Portability· 代理的记忆能否在模型升级中幸存？
Model upgrades are routine; memory migrations are not. An agent can keep the same memory s…
领域：cs.AI作者：Ankit Goyal、Jaideep Ray
论文精选 60已解读
Technical Manual for a Toolkit for Measuring Contextual Individuation in Transformer Language Models· 在变压器语言模型中测量上下文 individuation 的工具箱技术手册
A transformer language model assigns a single, context-independent vector to a word type a…
领域：cs.AI作者：José Luciano Verçosa Marques、Frederico Jorge Heitmann、Daniel Omar Perez
论文精选 60已解读
LLM-Driven Algorithm Design for Quantum Circuit Synthesis based on Binary Decision Diagrams
Quantum circuits are central to implementing quantum algorithms on quantum devices, where …
领域：cs.AI作者：Yoonju Sim、Federico Berto、Chuanbo Hua
论文精选 60已解读
Large Language Models for HVAC Operations in Building Energy Systems: A Critical Review of Methods, Applications, and Deployment Readiness· 大型语言模型在建筑能源系统中的应用
Building automation systems generate rich sensor data yet remain insight-poor because hete…
领域：cs.AI作者：Alexander Neubauer、Tianzhen Hong、Han Li
论文精选 82
LexFlip: A Dissociation Diagnostic for Legal Meaning Preservation Metrics
Does a simplified legal clause still say what the original said? The checks in current use…
领域：cs.CL作者：Gaurab Baral
论文精选 60已解读
RISE: Recursive Improvement via Self-Extrapolating Policy Distillation· RISE：通过自外推策略蒸馏递归改进
On-policy distillation (OPD) provides dense, per-token supervision for language model post…
领域：cs.AI作者：Yang Li、Semih Yavuz、Shafiq Joty
论文精选 85
Beyond Aggregate Scores: Behavioral Correctness Assumptions for Assessing Reference-Based Automatic Evaluation Methods· 超越总分：行为正确性假设
Automated reference-based evaluation methods play a critical role in assessing natural lan…
领域：cs.AI作者：Maria Mahbub、Ashley Rice、Michael R. Munroe
论文精选 83
GUT: Quantifying and Optimizing the Reasoning Uncertainty of LLMs via Graph Complexity· 量化和优化大模型推理不确定性
Recent years have witnessed great advances in the reasoning ability of Large Language Mode…
领域：cs.AI作者：Shuang Liang、Xin-Yu Hu、Xiang-Jun Ou
论文精选 85
Testing Interchangeability in LLM Agent Teams· 大模型智能体团队互换性测试
Production multi-agent systems replace agents constantly, on the assumption that an agent …
领域：cs.AI作者：Jianxin Gao、Tianyi Yu、Linna Deng
论文精选 85
Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference· 不要放弃Dropout：优化层稀疏性以提高LLM训练和推理效率
Layer dropout (a.k.a. stochastic depth) has been shown to enable faster training, higher a…
领域：cs.AI作者：Mostafa Elhoushi、Alex Pretko、Nolan Dey
论文精选 85
AI for Computational Design Science: A Responsible Human-AI Framework and Case Study on Short-Form Video Safety Surveillance· AI在计算设计科学中的应用框架与案例研究
Artificial intelligence (AI) is transforming not only what information systems researchers…
领域：cs.AI作者：Wenli Zhang、Jiaheng Xie、Zhihe Pan
论文精选 75
Self-Supervised Lexical Representation Learning for Fast, Large-Scale Phylogenetic Inference· 自监督词表示学习
Computational phylogenetics has become an essential tool in historical linguistics, yet it…
领域：cs.CL作者：Tim Wientzek
论文精选 85
Trace2Tower: Transition-Aware EigenTrace Induction of Multi-Level Skills for LLM Agents· Trace2Tower：面向转换的多级技能EigenTrace归纳
Large language model agents increasingly rely on execution traces to master complex intera…
领域：cs.AI作者：Jiazheng Sun、Boyu Yang、Binhao Yuan
论文精选 83
Commonsense Reasoning in Computer Vision: Foundations, Recent Advancements, and Future Directions· 计算机视觉中的常识推理：基础、进展与未来方向
Commonsense reasoning in computer vision encompasses integrating visual data and contextua…
领域：cs.AI作者：Bahar Uddin Mahmud、Sumit Barua、Guan Yue Hong
论文精选 85
A Unified Physics-Aware Quantum Machine Learning Framework across Power GaN HEMTs and Logic Nanowire FETs: Predicting Unseen Process Splits and Held-O...· 跨GaN HEMT和NWFET的物理感知量子机器学习框架
We present a unified reinforcement-learning (RL) framework that discovers compact parametr…
领域：cs.AI作者：Rushat Rai、Yun-Yuan Wang、Autsada Kakaen
论文精选 82
Do LLMs Exhibit Coherent Knowledge Structures in Mathematical Reasoning? A Perspective from Knowledge Space Theory· 大模型在数学推理中是否展现连贯的知识结构？
Human knowledge is inherently structured and interdependent: mastery of a concept requires…
领域：cs.AI作者：Peng Cui、Heejin Do、Mrinmaya Sachan
论文精选 82
A Verifier-Guided Explainable Reasoning Framework with Gold-Anchored QLoRA, Task-Aware Mixture-of-Experts, and Group-Relative RLVR· 基于验证器的可解释推理框架
Large language models (LLMs) show strong reasoning ability, but their explanations can rem…
领域：cs.CL作者：Thi Kim Trang Vo、Nam Tien Le、Thi Kim Nguyet Vo
论文精选 82
Can Large Language Models Anticipate Behavioral Responses to Social Policies? A Case of Pension Enrollment Prediction among China's Flexible Workers· 用大模型预测养老金参保行为
Assessing the impacts of social policy changes is a widely acknowledged challenge for poli…
领域：cs.CL作者：Yumiao Li、Peixin Liu、Donglin Di
GitHub 热门开源项目
开源项目精选 65
aimeoa/hanshuang-codex
针对GPT 5.6,V4flash的破甲实验项目
⭐ 38 stars语言 Python
开源项目精选 82
michailfragkiskos/EdgeMosaic
Rapid on-device AI evaluation lab: manifest validation, benchmark plans, latency/memory/th…
⭐ 28 stars语言 Python
开源项目精选 85
truespar/paddock· GPU 模型推理服务器
Native Rust inference server for open models on NVIDIA GPUs. OpenAI- and Anthropic-compati…
⭐ 62 stars语言 Rust
开源项目精选 82
naderelewa/Product-to-Prod· AI 产品管理插件
AI product management skills and plugin for Claude Code, Cowork, Codex & other AI agents: …
⭐ 41 stars语言 Shell
开源项目精选 80
horry0214/SideAsk· SideAsk：AI旁问层
AI side-question layer — ask aside, stay on track.
⭐ 44 stars语言 JavaScript
开源项目精选 85
arjun988/promptModel· PromptForge
PromptForge scores LLM prompts across seven quality dimensions, then rewrites weak prompts…
⭐ 33 stars语言 Jupyter Notebook
精选 AI 行业资讯
行业资讯精选 82
Mistral raises €3B as sovereign AI becomes big business· Mistral完成30亿欧元融资
The French AI lab has raised €3 billion at a €21 billion valuation in a Series D round led…
行业资讯精选 75
The Work Now Within Reach· 现在触手可及的工作
Explore how more capable, affordable AI can expand the work people and businesses can acco…
行业资讯精选 65
华为Mate XT 2 非凡大师发布：刘德华现身发布会再叙非凡精神
9月7日，HarmonyOS 7 | HUAWEI Mate XT 2 及全场景新品发布会举行，华为Mate XT 2 非凡大师正式亮相。在发布会上，除了产品的展示，华为Mate …
行业资讯精选 81
中国品牌出海的「清洁机器人答案」
2026年9月4日，柏林，IFA展馆。 9号馆里，一池碧水被搬进展厅。池子里游动的不是鱼，是一台正在自主巡航的泳池清洁机器人——这是石头科技全球首发的RockAqua P1。正对着…
行业资讯精选 82
CPU加速AI普及，GPU首次加AI，Arm为何把NPU留给伙伴？
个人AI的实现，到底需要一套怎样的计算底座？当手机开始理解用户意图、调用工具并连续完成任务，计算平台要支撑的已不只是一次模型推理，而是完整的个人AI体验。 <img src="ht…
行业资讯精选 83
百度搭子全面接入小度硬件，百度智能体进驻家庭空间
9月8日，在北京举行的百度AI Day小度新品发布会上，小度宣布超能小度完成智能体化升级，并发布多款全新的家庭场景智能体应用。百度搭子作为小度智能体能力的底座，全面落地小度智能屏、…
行业资讯精选 82
“ONE FOR ALL”背后，帕西尼正在构建怎样的物理AI闭环？
过去一个月，帕西尼（PaXini AI）的战略进程明显提速：发布搭载GEN4 FUSE真6D触觉感知芯片的PX6AX GEN4产品矩阵；北京总部落地，形成北京战略研发与深圳制造交付…
行业资讯精选 82
基元律动发布模型NeoHorse，探索Harness驱动的RSI路径
在一次项目排期测试中，一个4B基础模型找到了工作目录中的文件，却漏读了一封包含最新依赖约束的邮件。它按照过时信息生成计划，并把文件写到了错误位置。 这一案例来自基元律动（Token…
行业资讯精选 63
感谢用GPT-6搓3D的！燃烧自己的token换来一次全体重置· GPT-6生成3D模型
Blender也不用学了
行业资讯精选 82
硅谷AI独角兽转投阿里千问：Perplexity基于Qwen3.8打造本地Agent
9月8日消息，据美国科技媒体Siliconangle报道，硅谷AI独角兽公司Perplexity新推出一款本地Agent产品Portable Computer，采用的正是阿里最新开…
行业资讯精选 75
现场围观金融AI决赛，大厂挑人的逻辑我悟了
百万奖金、大厂直通、VC跟投
行业资讯精选 82
深度智控获宁德时代、沙特阿美战投等重磅加码，加速打造物理AI时代算力与能源底座
近日，物理AI企业深度智控（DeepCtrls）完成新一轮B+轮数亿元融资。
行业资讯精选 65
深入马来西亚AI现场！WAIC CONNECT MALAYSIA首日亮点全速递
从看市场，到见场景；从认识伙伴，到寻找合作。
行业资讯精选 82
王云鹤创业后交出首个模型
把多模型执行经验用到了模型训练
行业资讯精选 65
IFA观察：智能IoT进入「精耕」时代，萤石的技术储备开始兑现
<span mpa-font-style="mtq03ewv1w5k" style="font-size: 15px; color: rgba(0, 0, 0, 0.9); fon…
行业资讯精选 80
OpenAI expands initiatives to support journalism from classrooms to newsrooms· OpenAI 扩大新闻业支持计划
OpenAI is expanding support for journalism with tools, training, and partnerships for stud…
行业资讯精选 82
Opaque recurrence, and other AI terms that you should probably know· 你应该知道的AI术语
The rise of AI has brought an avalanche of new terms and slang. Here is a glossary with de…
YouTube 热门 AI 视频
AI 视频精选 80
Building Europe's AI Stack — Inside the Barcelona Supercomputing Center· 构建欧洲AI堆栈 — 走进巴塞罗那超算中心
YouTube · NVIDIA
📺 NVIDIA▶ 240 次观看
AI 视频精选 85
GPT-6 Astra Changes Everything· GPT-6 星系规模改变一切
YouTube · Two Minute Papers
📺 Two Minute Papers▶ 8.7k 次观看
AI 视频精选 69
GPT-6 Astra with Tom Krcha
YouTube · OpenAI
📺 OpenAI▶ 714 次观看