从参数化复杂度角度揭示了ReLU神经网络验证的困难性，为理解其计算极限提供新视角
arXiv:2509.22849v2 Announce Type: replace-cross Abstract: Neural networks with ReLU activations are a widely used model in machine learning. It is thu…
理论证明使用次指数尾随机梯度的高精度对数凹采样可实现多对数复杂度，与凸优化形成显著对比
arXiv:2602.14342v2 Announce Type: replace-cross Abstract: We show that high-accuracy guarantees for log-concave sampling -- that is, iteration and que…
首次给出非对数凹分布采样相对Fisher信息保证的查询复杂度，基于近端采样器与限制高斯oracle，是采样理论的重要进展
arXiv:2605.15859v1 Announce Type: cross Abstract: We study the query complexity of obtaining a relative Fisher information guarantee for sampling from…
GitHub创新图谱数据如何跨境揭示国家数字复杂度，研究者新视角。
Researchers share in an interview how they used GitHub data to predict GDP, inequality, and emissions in ways that traditional economic data misses, a…
这篇论文精准揭示了LLM在优化建模中的能力边界：纯文本推理随复杂度飙升鲁棒性崩溃，集成工具能算数却守不住全局约束，而自动公式化约束成为当前SOTA的核心瓶颈。一张结构化的路线图，指明下一代推理模型必须攻克的关卡。
arXiv:2601.19924v2 Announce Type: replace-cross Abstract: We investigate the capabilities and scalability of Large Language Models (LLMs) in optimizat…