1
Benign Overfitting Does Not Occur in Diffusion Models
解读扩散模型为何不会“良性过拟合”，帮你避开训练中的隐性地雷
arXiv:2607.02671v1 Announce Type: cross Abstract: Benign overfitting and double descent have come to shape our understanding of generalization in deep…
2
Tuning without Peeking: Provable Generalization Bounds and Robust LLM Post-Training
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…
3
Multi-LCB: Extending LiveCodeBench to Multiple Programming Languages
24个LLM大模型在多语言编程基准中暴露Python过拟合与语言污染，新基准Multi-LCB揭示显著性能差异
arXiv:2606.20517v1 Announce Type: new Abstract: LiveCodeBench (LCB) has recently become a widely adopted benchmark for evaluating large language model…
4
Effective Training Principles of Physical Reservoirs
破解光学储层计算过拟合难题，提出高效训练新原则
arXiv:2606.10130v1 Announce Type: cross Abstract: Reservoir computers benefit from the inherent complexity of optical phenomena, which provide rich, o…
5
PINNs Failure Modes are Overfitting
揭示物理信息神经网络（PINNs）失败的根本原因——过拟合，为改进PINN训练提供新视角
arXiv:2605.30910v1 Announce Type: new Abstract: Physics-Informed Neural Networks (PINNs) are a common class of machine learning-based partial differen…
6
A Principled Self-Referenced Early Stopping Approach for Deep Image Prior
提出一种无需额外数据的自参考早停方法，有效解决Deep Image Prior中的过拟合与退化问题，提升图像重建质量。
arXiv:2605.25299v1 Announce Type: cross Abstract: Recently, Deep Image Prior (DIP) has demonstrated strong capabilities for solving inverse imaging pr…
7
A Rigorous, Tractable Measure of Model Complexity
提出一种严谨且可计算的模型复杂度度量方法，为深度学习理论分析提供新工具
arXiv:2605.21167v1 Announce Type: cross Abstract: An accurate assessment of a model's complexity is crucial for topics such as interpretation, general…