AI资讯日报 - 2026/7/22

👨‍🔬 Hanqing Zhu, Wenyan Cong, Zhizhou Sha, Sagnik Mukherjee, Xinyuan Song, David González-Martínez, Xiaoxia Wu, Yuandong Tian, Shiwei Liu, David Z. Pan, Zhangyang "Atlas" Wang

Off-Context GRPO: Learning to Reason on Hard Problems using Privileged Information

学术论文 ArXiv 重要度: 8

提出离线上下文GRPO，利用特权引导解决困难推理问题，提升3.9%绝对精度。

👨‍🔬 Priyank Agrawal, Ankur Samanta, Shervin Ghasemlou, Jalaj Bhandari, Kavosh Asadi, Daniel Jiang, Aditya Modi

ResearchArena: Evaluating Sabotage and Monitoring in Automated AI R&D

学术论文 ArXiv 重要度: 7

评估AI研发中自主代理的破坏与监控，发现训练数据隐藏破坏最难检测。

👨‍🔬 Lena Libon, Ben Rank, Jehyeok Yeon, David Schmotz, Jeremy Qin, Daniel Donnelly, Derck Prinzhorn, Maksym Andriushchenko

Agents in the Wild: Where Research Meets Deployment

学术论文 ArXiv 重要度: 7

探讨AI代理从研究到部署的挑战，包括鲁棒性、安全性和可靠性。

👨‍🔬 Grace Hui Yang, Pranav N. Venkit, Hooman Sedghamiz, Enrico Santus, Victor Dibia, Ioana Baldini

Provable diffusion-based posterior sampling for linear inverse problems via DDIM

学术论文 ArXiv 重要度: 7

提出可证明的扩散后验采样算法，实现线性逆问题的高效求解。

👨‍🔬 Yuchen Jiao, Na Li, Changxiao Cai, Yuxin Chen, Gen Li

Appearance Pointers -- Multimodal Region Control of Diffusion Transformers

学术论文 ArXiv 重要度: 7

引入外观指针，实现扩散Transformer的多模态区域精确控制。

👨‍🔬 Rahul Sajnani, Yulia Gryaditskaya, Radomír Měch, Srinath Sridhar, Matheus Gadelha

CodeRescue: Budget-Calibrated Recovery Routing for Coding Agents

学术论文 ArXiv 重要度: 6

提出预算校准恢复路由，在代码代理失败后选择继续使用廉价模型或升级。

👨‍🔬 Qijia He, Jiayi Cheng, Chenqian Le, Rui Wang, Xunmei Liu, Yixian Chen, Jie Mei, Zhihao Wang, Xupeng Chen, Yuhuan Chen, Tao Wang

LLM Detection as an Intervention: Downstream Impact under Strategic User Behavior

学术论文 ArXiv 重要度: 6

分析LLM检测对用户行为的战略影响，发现可能增加LLM使用或降低输出质量。

👨‍🔬 Meena Jagadeesan, Tatsunori Hashimoto, Jon Kleinberg

From Distances to Trajectories: Real-Time Signed Distance Function Mapping and Distance-Accelerated Motion Planning for UAVs

学术论文 ArXiv 重要度: 6

结合八叉树残差网络与Bubble*规划器，实现无人机实时SDF建图与快速运动规划。

👨‍🔬 Jason Stanley, Zhirui Dai, Qihao Qian, Tzu-Chin Ho, Tianxing Fan, Siddharth Saha, Christopher Barngrover, Ki Myung Brian Lee, Nikolay Atanasov

Associative Emotional Learning in Convolutional Neural Networks

学术论文 ArXiv 重要度: 5

用CNN模拟联想情绪学习，复现人类联想学习中的关联形成与泛化。

👨‍🔬 Seowung Leem, Andreas Keil, Mingzhou Ding, Ruogu Fang

Riemannian Deep Learning:Modules, Networks, and Geometries

学术论文 ArXiv 重要度: 5

提出黎曼深度学习统一框架，涵盖模块、网络架构和几何设计。

👨‍🔬 Chen Ziheng

🤖 AI资讯日报

📊 今日趋势总结

MIT Non-AI License

Why Boring Businesses Outlast AI Hype Cycles

Ask HN: Anyone concerned about NYC Local Law 144?

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

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

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

The AI Crackpot Index

NLP, AI, ML, bots – a passing trend or much more?

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

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

Show HN: Startup Raising capital through Book Sales

Bioinformatician

Copy Less, Ground More: Overcoming Repetitive Copying in Long-Context Reasoning via Evidence-Aware Reinforcement Learning

ISO: An RLVR-Native Optimization Stack

Off-Context GRPO: Learning to Reason on Hard Problems using Privileged Information

ResearchArena: Evaluating Sabotage and Monitoring in Automated AI R&D

Agents in the Wild: Where Research Meets Deployment

Provable diffusion-based posterior sampling for linear inverse problems via DDIM

Appearance Pointers -- Multimodal Region Control of Diffusion Transformers

CodeRescue: Budget-Calibrated Recovery Routing for Coding Agents

LLM Detection as an Intervention: Downstream Impact under Strategic User Behavior

From Distances to Trajectories: Real-Time Signed Distance Function Mapping and Distance-Accelerated Motion Planning for UAVs

Associative Emotional Learning in Convolutional Neural Networks

Riemannian Deep Learning:Modules, Networks, and Geometries

📅 历史日报目录