一份用信赖域框架重新审视自适应矩估计的数学研究，适合想深挖优化器收敛原理的读者。
arXiv:2608.04026v1 Announce Type: cross Abstract: In this paper, we develop a trust-region framework for understanding the behavior of adaptive moment…
一作质问持续学习何时真正需要「学习」，揭示任务无关场景下模型无需更新即可泛化，引发对学习本质的再思考。
arXiv:2607.07847v1 Announce Type: new Abstract: As large language models (LLMs) become increasingly capable, the next question is how can we enable mo…
将带噪梯度下降理想化为过阻尼朗之万动力学，量化模型参数落入危险区域的概率，为训练安全提供理论保证。
arXiv:2607.07538v1 Announce Type: new Abstract: Training a model with noisy gradient descent can be idealized as overdamped Langevin dynamics on the l…
用可列表解码编码推动Boosting理论，交叉编码与学习的新视角
arXiv:2607.05791v1 Announce Type: cross Abstract: Boosting is a fundamental technique for generically improving the accuracy of learning algorithms (S…
一项新研究揭示Transformer在训练中会收敛到不变的算法核心，为理解模型行为与电路机制提供新视角。
arXiv:2602.22600v2 Announce Type: replace-cross Abstract: Training selects for behavior, not circuitry: many weight configurations can implement the s…
生成式AI模型是否会概率性地“复制”训练数据？这篇论文从理论上剖析了模型记忆与版权边界。
Article URL: https://download.ssrn.com/2026/7/6/7067878.pdf?response-content-disposition=inline&X-Amz-Security-Token=IQoJb3JpZ2luX2VjELH%2F%2F%2F%…
从布朗运动路径提取签名特征，理论证明其具备全局通用逼近能力，为序列建模打开新思路。
arXiv:2512.16396v2 Announce Type: replace-cross Abstract: We establish $L^p$-universal approximation theorems for general path-dependent and non-antic…
从学习理论视角解析模型崩溃，重放机制能否成为破局关键？ICML 2026 新作给出严谨答案。
arXiv:2603.11784v2 Announce Type: replace Abstract: As scaling laws push the training of frontier large language models (LLMs) toward ever-growing dat…
从自然语言统计规律出发，揭示神经缩放定律的数学根源，为理解大模型能力增长提供理论基石。
arXiv:2602.07488v3 Announce Type: replace-cross Abstract: Despite the fact that experimental neural scaling laws have substantially guided empirical p…
颠覆知识蒸馏常规认知：预训练表征只有等价类意义，匹配坐标是伪命题
arXiv:2607.03572v1 Announce Type: cross Abstract: Knowledge distillation is usually framed as a choice of what to match in the teacher - its logits, h…
从逼近到涌现，为深度学习提供统一理论框架，AI研究者必读的硬核前沿。
arXiv:2607.01311v1 Announce Type: new Abstract: Deep learning has outgrown any single mathematical explanation. From Approximation to Emergence develo…
大模型先验如何提升程序搜索中的经验风险最小化？理论+方法前沿新作
arXiv:2510.14331v3 Announce Type: replace Abstract: We study program-learning methods that are efficient in both samples and computation. Classical le…
从理论层面剖析持续学习如何抵御数据投毒攻击，为鲁棒AI提供新视角。
arXiv:2606.29841v1 Announce Type: new Abstract: Continual learning (CL), where a model is trained on a sequence of data tasks, is increasingly being a…
凸学习与非仿射聚合看似美妙结合，实则暗藏“危险关系”，这篇论文揭示了潜藏的理论风险。
arXiv:2606.28123v1 Announce Type: new Abstract: Last-iterate convergence and generalization guarantees in first-order convex learning hinge on the mon…
重访等价查询学习，理论深度突破，机器学习研究者必读的前沿论文。
arXiv:2604.04535v2 Announce Type: replace Abstract: Modern machine learning systems, such as generative models and recommendation systems, often evolv…
探索测度近似的新结构化方法，为机器学习和概率论提供理论工具
arXiv:2310.09149v3 Announce Type: replace-cross Abstract: We study the approximation of probability measures in the Wasserstein-$p$ distance by struct…
提出“无窥视调优”方法，为大模型后训练提供可证明的泛化界限与鲁棒性保障。
arXiv:2507.01752v4 Announce Type: replace-cross Abstract: Gradient-based optimization is the workhorse of deep learning, offering efficient and scalab…
证明EML树具备通用逼近能力，为混合机器学习架构奠定理论基础。
arXiv:2606.23179v1 Announce Type: new Abstract: The recently introduced EML (Exp-Minus-Log) function acts as continuous analogue of NAND gates, provid…
将形式化验证与PAC-Bayesian理论结合，为过程奖励模型提供可证实的泛化边界，是AI安全领域的硬核进展。
arXiv:2606.20740v1 Announce Type: cross Abstract: Process Reward Models (PRMs) provide step-level verification for Large Language Model (LLM) reasonin…
从统计视角剖析训练与泛化机制，为机器学习理论提供严谨数学支撑。
arXiv:2606.20299v1 Announce Type: cross Abstract: Deep learning has managed to evade numerous intuitions from classical statistics to achieve unpreced…