1
Optimization with SpotOptim
SpotOptim优化器新工作论文，十八页篇幅系统介绍方法思路，技术细节与理论分析兼备，值得算法研究者关注。
arXiv:2604.13672v2 Announce Type: replace Abstract: The spotoptim package implements surrogate-model-based optimization of expensive black-box functio…
2
Deep Neural Networks Inspired by Differential Equations
从微分方程视角解读深度神经网络，帮你打通数学与AI的任督二脉，理论进阶必读。
arXiv:2510.09685v2 Announce Type: replace-cross Abstract: Deep learning has become a pivotal technology in fields such as computer vision, scientific …
3
Delegation and Verification Under AI
聚焦AI系统下的任务委托与验证机制，ICML 2026收录的前沿理论，值得研究者关注。
arXiv:2603.02961v2 Announce Type: replace-cross Abstract: As AI systems enter institutional workflows, workers must decide whether to delegate task ex…
4
Understanding Latent Diffusability via Fisher Geometry
读懂潜在扩散模型Fisher几何视角，揭示扩散可区分性原理，助力生成模型优化与理解
arXiv:2604.02751v2 Announce Type: replace Abstract: Diffusion models often degrade in latent spaces, yet the formal causes remain poorly understood. W…
5
AI Contagion in Social Networks
从经济学视角建模AI在社交网络中的传播机制，解读信息扩散的新规律。
arXiv:2606.15206v1 Announce Type: cross Abstract: We study how artificial intelligence (AI) interacts with social communication networks to shape the …
6
Chain Of Thought Compression: A Theoretical Analysis
深入剖析思维链压缩的理论基础，揭示大模型推理效率提升的关键机制。
arXiv:2601.21576v2 Announce Type: replace Abstract: Chain-of-Thought (CoT) has unlocked advanced reasoning abilities of Large Language Models (LLMs) w…
7
Commutator-Induced Uncertainty in VAEs
VAE新视角：用交换子数学工具量化模型不确定性，理论严谨有深度
arXiv:2605.23449v1 Announce Type: new Abstract: Variational autoencoders (VAEs) often struggle to represent non-commutative structure in learned laten…