Why is SAM Robust to Label Noise?

C Baek, Z Kolter, A Raghunathan - arXiv preprint arXiv:2405.03676, 2024 - arxiv.org
Sharpness-Aware Minimization (SAM) is most known for achieving state-of the-art
performances on natural image and language tasks. However, its most pronounced …

Optimization dependent generalization bound for ReLU networks based on sensitivity in the tangent bundle

D Rácz, M Petreczky, A Csertán, B Daróczy - arXiv preprint arXiv …, 2023 - arxiv.org
Recent advances in deep learning have given us some very promising results on the
generalization ability of deep neural networks, however literature still lacks a comprehensive …