牛哥精选 · 所有

📋 全部 🤖 AI·大模型 ⚡ 效率工具 📝 深度技术 🚀 产品观察 💰 商业科技 🔓 开源项目 🎨 设计创意 📖 阅读推荐 🏷 资源合集 🌱 成长效率

📝 深度技术 arXiv 机器学习 2026-06-16

Tail-Shape Estimation in LLM Evaluation Is Fragile: A Protocol for Diagnosing False Positives

揭示LLM评估中尾部分布估计的脆弱性，并提出诊断假阳性的新协议，对模型评测有重要参考价值。

arXiv:2606.16511v1 Announce Type: new Abstract: Recent work motivates moving large language model (LLM) evaluation from mean-based to tail-aware metri…

llm评估尾形状估计假阳性诊断协议机器学习

🤖 AI·大模型 Dev.to 2026-06-16

AI Detectors Are Failing Students — Here's What Universities Actually Know

AI检测器宣称1%误报率，但实际可能误判750名学生，这篇文章揭露大学面临的真实困境。

The detector said I cheated. I wrote every word myself. That's the opening line of a Reddit post from a public health student at the University at Buf…

ai检测器假阳性学生教育 turnitin

🤖 AI·大模型 arXiv AI 2026-05-29

Physics Is All You Need? A Case Study in Physicist-Supervised AI Development of Scientific Software

AI自主修复10个bug，物理学家用领域知识再补2个，但假阳性修复暴露了根因误判的经典短板

arXiv:2605.30353v1 Announce Type: new Abstract: Are AI agents tools, co-authors, or researchers? We present a quantified case study ($N=1$): a physici…

ai开发科学软件物理学家监督 bug修复假阳性

📅 日期

2026-05-20 2026-05-19

🐂 牛哥精选

Tail-Shape Estimation in LLM Evaluation Is Fragile: A Protocol for Diagnosing False Positives

AI Detectors Are Failing Students — Here's What Universities Actually Know

Physics Is All You Need? A Case Study in Physicist-Supervised AI Development of Scientific Software

📅 日期