AI资讯日报 - 2026/7/31

👨‍🔬 Xiangning Lin, Shenzhe Zhu, Shu Yang, Zhenyu Zhang, Haoqian Zhang, Yipeng Zhao, Chengxuan Qian, Tianwei Wang, Ziheng Zhang, Zhenlong Yuan, Dingcheng Wang, Juncheng Wu, Yuan Si, Jiaxin Liu, Baolong Bi, Robert Mahari, Tobin South, Dazza Greenwood, Zexue He, Rishi Bommasani, Sophia Kazinnik, Andreas Haupt, Samuele Marro, Erik Brynjolfsson, Alex Pentland, Jiaxin Pei

OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

学术论文 ArXiv 重要度: 9

提出OSReward基准评估VLM裁判，发现现有模型存在系统性宽松偏差，并训练低成本的开放奖励模型OS-Shepherd。

👨‍🔬 Qiushi Sun, Kanzhi Cheng, Yian Wang, Bowen Yang, Hang Yan, Liheng Chen, Fangzhi Xu, Zichen Ding, Nuo Chen, Jialin Cao, Xingdong Gong, Zehao Li, Kaiming Jin, Xinfeng Yuan, Zhoumianze Liu, Jingyang Gong, Zhangyue Yin, Jiahui Gao, Zhiyong Wu, Tianbao Xie, Jianbing Zhang, Ben Kao, Lingpeng Kong

AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

学术论文 ArXiv 重要度: 8

构建以主张为中心的化学文献检索基础设施，将论文转化为带来源的原子化主张，支持跨论文综合。

👨‍🔬 Bing Yan, Gregory Wolfe, Stefano Martiniani, Kyunghyun Cho

PAIChecker: Uncovering and Checking PR-Issue Misalignment in SWE-Bench-Like Benchmarks

学术论文 ArXiv 重要度: 8

系统研究SWE-bench基准中PR-Issue错配问题，提出多智能体系统PAIChecker进行检测，准确率高。

👨‍🔬 Manyi Wang, Junjielong Xu, Pinjia He

Sample More, Reflect Less: Self-Refine and Reflexion Lose to Repeated Sampling at Equal Token Cost, from 1.5B to 7B

学术论文 ArXiv 重要度: 8

实验表明在相同token预算下，重复采样优于自我精炼和反思方法，质疑此类方法有效性。

👨‍🔬 Iliya Mirzaei

Rethinking Inference-Time Scaling in Local Computer-Use Agents: Failure Modes and Compute Tradeoffs

学术论文 ArXiv 重要度: 8

系统研究本地计算机使用智能体的推理时间扩展，发现额外计算常收益递减并改变失败模式。

👨‍🔬 Woongkyu Lee, Jungwook Choi

ReToken: One Token to Improve Vision-Language Models for Visual Retrieval

学术论文 ArXiv 重要度: 7

提出ReToken，通过可学习标记从预填充视觉KV缓存中选择相关视觉token，提升长视觉上下文的检索性能。

👨‍🔬 Yao Xiao, Reuben Tan, Zhen Zhu, Yuqun Wu, Jianfeng Gao, Derek Hoiem

PAC-MAN: Perception-Aware CBF-RL for Whole-Body Safety in Humanoid Dodgeball

学术论文 ArXiv 重要度: 7

提出感知感知的CBF-RL框架，用于人形机器人躲避球任务，在真实机器人上实现95%的成功率。

👨‍🔬 Lizhi Yang, Junheng Li, Aaron D. Ames

DualG-MRAG: Decoupling Macro-Reasoning and Micro-Matching for Multimodal Retrieval-Augmented Generation

学术论文 ArXiv 重要度: 7

提出双图框架解耦宏观推理与微观匹配，用于多模态RAG，提升复杂多跳推理性能。

👨‍🔬 Jiacheng Tao, Qingyun Sun, Haonan Yuan, Ziwei Zhang, Jianxin Li

APO: Unsupervised Atomic Policy Optimization for 3D Structure Prediction of Atomic Systems

学术论文 ArXiv 重要度: 7

提出无监督原子策略优化框架，无需真实坐标即可预测原子系统3D结构，超越监督基线。

👨‍🔬 Shentong Mo, Yatao Bian

Learning to Trace Seiberg Dualities

学术论文 ArXiv 重要度: 6

利用机器学习方法学习Seiberg对偶，建立quiver突变工具，并比较不同网络架构的性能。

👨‍🔬 Jonathan J. Heckman, Shani Meynet, Alessandro Mininno, Gary Shiu

Algorithms for Structured Elections under Thiele Voting Rules

学术论文 ArXiv 重要度: 5

研究Thiele投票规则下获胜者确定问题的计算复杂性，在受限域上提出FPT算法。

👨‍🔬 Alexandra Lassota, Krzysztof Sornat

🤖 AI资讯日报

📊 今日趋势总结

MIT Non-AI License

Ask HN: Anyone concerned about NYC Local Law 144?

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

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

Why Boring Businesses Outlast AI Hype Cycles

The AI Crackpot Index

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

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

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

Show HN: Startup Raising capital through Book Sales

Bioinformatician

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

huggingface/speech-to-speech

AISPA: User-Centric System Prompt Auditing for Large Language Model Applications

OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models

AskChem: Claim-Centered Infrastructure for Chemistry Literature Synthesis

PAIChecker: Uncovering and Checking PR-Issue Misalignment in SWE-Bench-Like Benchmarks

Sample More, Reflect Less: Self-Refine and Reflexion Lose to Repeated Sampling at Equal Token Cost, from 1.5B to 7B

Rethinking Inference-Time Scaling in Local Computer-Use Agents: Failure Modes and Compute Tradeoffs

ReToken: One Token to Improve Vision-Language Models for Visual Retrieval

PAC-MAN: Perception-Aware CBF-RL for Whole-Body Safety in Humanoid Dodgeball

DualG-MRAG: Decoupling Macro-Reasoning and Micro-Matching for Multimodal Retrieval-Augmented Generation

APO: Unsupervised Atomic Policy Optimization for 3D Structure Prediction of Atomic Systems

Learning to Trace Seiberg Dualities

Algorithms for Structured Elections under Thiele Voting Rules

📅 历史日报目录