AI资讯日报 - 2026/8/10

👨‍🔬 Daniel Armstrong, Xuan-Vu Nguyen, Octavian Susanu, Gabriel Gibberd, Théo A. Neukomm, Taddäus Strunden, Dan Forster, Morgane Delattre, Shawn Teh, Clément Rols, John Federice, Hayden Leatherwood, M. Lavelle Barnes, Maarten R. Dobbelaere, Peter Wipf, Jon T. Njardarson, Jieping Zhu, Philippe Schwaller

SkillProx: Self-Evolving Agent Skills via Proximal Textual Gradient Descent

学术论文 ArXiv 重要度: 7

引入近端梯度下降框架，通过闭环诊断和效用感知优化，提升代理技能准确性。

👨‍🔬 Mingxuan Zheng, Yujin Zhou, Chuxue Cao, Boqin Yin, Yuyao Zhang, Jiapeng Sun, Shuaishuai Gong, Sirui Han, Yike Guo

CoinRAG: Contextualized Information Nugget KV Cache Reuse for Long-Context RAG

学术论文 ArXiv 重要度: 7

通过细粒度缓存重用优化长上下文RAG，降低延迟并提升答案质量。

👨‍🔬 Gyuwan Kim, Cheoneum Park, Tao Yang

Taxonomy-Driven Analysis of Open-Source AI Risk Mitigation Tools

学术论文 ArXiv 重要度: 7

通过分类法分析21个开源工具，揭示风险缓解工具分布不均，提出分层架构。

👨‍🔬 Afreen Alam, Evgenija Popchanovska, Ana Gjorgjevikj, Maryan Rizinski, Lubomir T. Chitkushev, Irena Vodenska, Dimitar Trajanov

SABRE: Scalable and Automated Benchmarking of VLMs under Stress

学术论文 ArXiv 重要度: 7

自动化生成压力测试，揭示视觉语言模型在反事实场景下的弱点。

👨‍🔬 Zixuan Lan, Luzhe Sun, Matthew R. Walter, Jiawei Zhou

Interaction Creates Dynamical AI Behavior Absent in Isolation

学术论文 ArXiv 重要度: 6

发现AI代理交互时会出现孤立时没有的动态行为，为物理学提供新视角。

👨‍🔬 Bella Xinrui Li, Frank Yingjie Huo, Neil F Johnson

Blast Radius

学术论文 ArXiv 重要度: 6

提出预测性记忆管理层，通过可逆驱逐降低令牌消耗，提升代理编码可持续性。

👨‍🔬 MY Pitsane, Hope Mogale

PsychoAgent: An Affect-Sensitive Cognitive Architecture for Conflict-Aware Memory in LLM Agents

学术论文 ArXiv 重要度: 6

提出情感敏感认知架构，整合情感记忆，提升冲突情境下的检索能力。

👨‍🔬 Mohammad Amanlou, Parham Abed Azad, Farbod Davoodi, Mostafa Masumi, Behnam Bahrak, Abdol-Hossein Vahabie

Post-Grokking Collapse at the Representation-Readout Interface in Muon-Trained Transformers

学术论文 ArXiv 重要度: 6

发现Muon优化器训练后出现泛化崩溃，并分析其表示-读出界面的原因。

👨‍🔬 Ali Janati, Kaoutar El Maghraoui, Andrei Kanavalau, Anass Belfatmi

🤖 AI资讯日报

📊 今日趋势总结

MIT Non-AI License

Ask HN: What's the pain using current AI algorithms?

Why Boring Businesses Outlast AI Hype Cycles

The AI Crackpot Index

Ask HN: Anyone concerned about NYC Local Law 144?

Ask HN: Is the rate of progress in AI exponential?

Ask HN: What would you read to learn about "artificial intelligence"?

The Next Bill Gates or Albert Einstein in AI “Chris Clark” – Yourobot

NLP, AI, ML, bots – a passing trend or much more? What's your take on this?

Show HN: Startup Raising capital through Book Sales

Common Lisp + Machine Learning Internship at Google (Mountain View, CA)

Bioinformatician

Fisher-R1: Training LLM Agents for Reliable Hypothesis Testing

Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits

CreativeInstruct: Scalably Teaching LLMs to Balance Quality, Creativity, and Diversity

Strategy-first synthesis planning for complex natural products

SkillProx: Self-Evolving Agent Skills via Proximal Textual Gradient Descent

CoinRAG: Contextualized Information Nugget KV Cache Reuse for Long-Context RAG

Taxonomy-Driven Analysis of Open-Source AI Risk Mitigation Tools

SABRE: Scalable and Automated Benchmarking of VLMs under Stress

Interaction Creates Dynamical AI Behavior Absent in Isolation

Blast Radius

PsychoAgent: An Affect-Sensitive Cognitive Architecture for Conflict-Aware Memory in LLM Agents

Post-Grokking Collapse at the Representation-Readout Interface in Muon-Trained Transformers

📅 历史日报目录