AI资讯日报 - 2026/6/5

👨‍🔬 Shangheng Du, Xiangchao Yan, Jinxin Shi, Zongsheng Cao, Shiyang Feng, Zichen Liang, Boyuan Sun, Tianshuo Peng, Yifan Zhou, Xin Li, Jie Zhou, Liang He, Bo Zhang, Lei Bai

Goedel-Architect: Streamlining Formal Theorem Proving with Blueprint Generation and Refinement

学术论文 ArXiv 重要度: 8

提出形式化定理证明框架，通过蓝图生成和细化实现高通过率，成本低。

👨‍🔬 Jui-Hui Chung, Ziyang Cai, Zihao Li, Qishuo Yin, Rohit Agarwal, Simon Park, Rodrigo Porto, Narutatsu Ri, Ziran Yang, Shange Tang, Xingyu Dang, Hongzhou Lin, Mengdi Wang, Danqi Chen, Chi Jin, Liam H Fowl, Sanjeev Arora

You Only Index Once: Cross-Layer Sparse Attention with Shared Routing

学术论文 ArXiv 重要度: 8

提出跨层稀疏注意力，共享路由索引，实现7.6倍解码加速。

👨‍🔬 Yutao Sun, Yanqi Zhang, Li Dong, Jianyong Wang, Furu Wei

Code2LoRA: Hypernetwork-Generated Adapters for Code Language Models under Software Evolution

学术论文 ArXiv 重要度: 7

提出超网络生成仓库特定LoRA适配器，无需推理开销，在代码任务上达到微调上界。

👨‍🔬 Liliana Hotsko, Yinxi Li, Yuntian Deng, Pengyu Nie

Operation-Guided Progressive Human-to-AI Text Transformation Benchmark for Multi-Granularity AI-Text Detection

学术论文 ArXiv 重要度: 7

构建渐进人-AI文本转换基准，揭示混合作者文本检测的非单调模式。

👨‍🔬 Sondos Mahmoud Bsharat, Jiacheng Liu, Xiaohan Zhao, Tianjun Yao, Xinyi Shang, Yi Tang, Jiacheng Cui, Ahmed Elhagry, Salwa K. Al Khatib, Hao Li, Salman Khan, Zhiqiang Shen

Pretraining Recurrent Networks without Recurrence

学术论文 ArXiv 重要度: 7

提出监督记忆训练，通过一步转换标签训练RNN，实现时间并行和长程依赖。

👨‍🔬 Akarsh Kumar, Phillip Isola

Self-Augmenting Retrieval for Diffusion Language Models

学术论文 ArXiv 重要度: 7

提出扩散语言模型的自增强检索框架，利用低置信度令牌引导检索，无需训练。

👨‍🔬 Paul Jünger, Justin Lovelace, Linxi Zhao, Dongyoung Go, Kilian Q. Weinberger

PC Layer: Polynomial Weight Preconditioning for Improving LLM Pre-Training

学术论文 ArXiv 重要度: 7

提出多项式预条件层改善权重条件数，提升LLM预训练效果，无推理开销。

👨‍🔬 Senmiao Wang, Tiantian Fang, Haoran Zhang, Yushun Zhang, Kunxiang Zhao, Alex Schwing, Ruoyu Sun

Regret Minimization with Adaptive Opponents in Repeated Games

学术论文 ArXiv 重要度: 6

提出重复策略遗憾指标，设计算法在自适应对手下最小化遗憾，可学习均衡。

👨‍🔬 Mingyang Liu, Asuman Ozdaglar, Tiancheng Yu, Kaiqing Zhang

RREDCoT: Segment-Level Reward Redistribution for Reasoning Models

学术论文 ArXiv 重要度: 6

提出推理模型奖励再分配方法，利用模型自身近似最优奖励，降低方差。

👨‍🔬 Mykyta Ielanskyi, Kajetan Schweighofer, Lukas Aichberger, Sepp Hochreiter

🤖 AI资讯日报

📊 今日趋势总结

MIT Non-AI License

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

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

Ask HN: Anyone concerned about NYC Local Law 144?

Why Boring Businesses Outlast AI Hype Cycles

The AI Crackpot Index

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

Show HN: Startup Raising capital through Book Sales

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

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

Bioinformatician

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

HANDOFF: Humanoid Agentic Task-Space Whole-Body Control via Distilled Complementary Teachers

TempoVLA: Learning Speed-Controllable Vision-Language-Action Policies

MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery

Goedel-Architect: Streamlining Formal Theorem Proving with Blueprint Generation and Refinement

You Only Index Once: Cross-Layer Sparse Attention with Shared Routing

Code2LoRA: Hypernetwork-Generated Adapters for Code Language Models under Software Evolution

Operation-Guided Progressive Human-to-AI Text Transformation Benchmark for Multi-Granularity AI-Text Detection

Pretraining Recurrent Networks without Recurrence

Self-Augmenting Retrieval for Diffusion Language Models

PC Layer: Polynomial Weight Preconditioning for Improving LLM Pre-Training

Regret Minimization with Adaptive Opponents in Repeated Games

RREDCoT: Segment-Level Reward Redistribution for Reasoning Models

📅 历史日报目录