News for 2026-10-02
193 items that day
- ·arXiv AI
From Proposal to Verified Effect: Praxa, an Evidence-Bound Harness for Governed AI Agent Execution
- ·arXiv AI
What Do Rationales Communicate? A Message-Intervention Study in Role-Specialized QA
- ·arXiv AI
Measuring the Microtask Eligibility Gap: When Is an Off-the-Shelf SLM Enough for an Agent Harness?
- ·arXiv AI
Characterizing a Configuration Where Inference-Time PRM-Pruned Fragment Grafting Is Inert: Evidence from Three Reasoning LMs
- ·arXiv AI
Gradient-Aligned Pair Selection for Personalized Preference Optimization
- ·arXiv AI
Scientific Agents: Evaluating Profession-Specific System Prompts on Scientific Tasks
- ·arXiv AI
Comedic Fool's Gold: Reward Exploits and Countermeasures in Conversational Humor
- ·arXiv AI
When Do Causal World Models Help Modular LLM Agents
- ·RadarAI
OpenAI 通报 100 多家第三方机构,称旗下 AI 智能体或曾试图绕过安全防护
- ·RadarAI
OpenAI dot 可创建宠物头像
- ·RadarAI
Broadcom 同意向 Anthropic 提供最高 420 亿美元贷款用于租赁其 AI 芯片
- ·RadarAI
AgentWorld 论文:多智能体协作不足三分之一有效
- ·RadarAI
贝恩报告:全球 AI 行业到 2031 年需创造 6 万亿美元年营收才能支撑数据中心投入
- ·RadarAI
Hacker News 投票:哪些 AI 挑战已经实现
- ·RadarAI
AI 伦理研究:DeepSeek V4-Flash 对男女一视同仁,Claude Sonnet 4.6 与 GPT-5.5 却“区别对待”
- ·RadarAI
Claude 助我做游戏跳跃动画编辑器
- ·RadarAI
AI 影视大考:开源项目落地,演员授权长剧来了
- ·RadarAI
消息称 Anthropic 寻求最早 11 月中旬启动 IPO,争取感恩节前挂牌
- ·RadarAI
MerchantBench 评测显示八款 LLM 长周期任务表现远逊人类,最好模型仅达人类 27.3%
- ·RadarAI
ChatGPT 付费账户明日全球重置
- ·RadarAI
Signal65 报告:高通骁龙 X2 Elite 在 CPU / AI 等方面优于英特尔酷睿 Ultra X7 358H
- ·RadarAI
Meta 用控制器调度长程智能体推理,ProgramBench 提升至 71.5%
- ·RadarAI
MiniMax 将亮相 AI Engineer NYC
- ·RadarAI
BestBlogs 早报:Claude 科研协作与模型路由降本 64%
- ·RadarAI
BestBlogs 早报:Claude 科研工具与软件工厂 ROI
- ·RadarAI
Volantis 获 8800 万美元 A 轮融资,用光子互连为大模型推理提速
- ·RadarAI
Google DeepMind 发布 SynthID Bio 为 AI 设计的蛋白质添加水印
- ·RadarAI
Volantis 融资 8800 万美元 A 轮,用光学技术突破内存瓶颈加速推理
- ·RadarAI
Karpathy 分享用 LLM 生成写作、网页和讲解视频的技巧,Yuchen Jin 盼其回归更新视频
- ·RadarAI
Google 首次轨道 AI 芯片试验确认在轨运行正常