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

🤖 AI·大模型 arXiv AI 2026-07-03

Who Gets the Reward & Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents

多LLM智能体如何公平分配奖励与责任？这篇论文提出评价对齐的训练信号，解决多智能体协作中的信用分配难题。

arXiv:2511.10687v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) in multi-agent systems (MAS) have shown promise for complex tas…

多智能体大语言模型训练信号信用分配评价对齐

🤖 AI·大模型 arXiv NLP 2026-06-02

Learning from Saturated Data: Signals Beyond Correctness for LLM Training

LLM训练新思路：从“饱和数据”中挖掘超越正确性的隐藏信号

arXiv:2606.01436v1 Announce Type: new Abstract: The growing capabilities of large language models (LLMs) have led to the saturation of many benchmarks…

llm训练饱和数据超越正确性训练信号 arxiv论文

📝 深度技术 arXiv AI 2026-06-01

Pull Requests as a Training Signal for Repo-Level Code Editing

被ICML 2026接收，提出用Pull Requests作为训练信号来提升仓库级别代码编辑能力，开辟AI编程新路径。

arXiv:2602.07457v2 Announce Type: replace-cross Abstract: Repository-level code editing requires models to understand complex dependencies and execute…

代码编辑 pull reque 训练信号仓库级别 icml 2026

📅 日期

2026-05-20 2026-05-19

🐂 牛哥精选

Who Gets the Reward & Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents

Learning from Saturated Data: Signals Beyond Correctness for LLM Training

Pull Requests as a Training Signal for Repo-Level Code Editing

📅 日期