Seguir
Jiancong Xiao
Título
Citado por
Citado por
Año
Stability analysis and generalization bounds of adversarial training
J Xiao, Y Fan, R Sun, J Wang, ZQ Luo
Advances in Neural Information Processing Systems 35, 15446-15459, 2022
282022
Adversarial rademacher complexity of deep neural networks
J Xiao, Y Fan, R Sun, ZQ Luo
arXiv preprint arXiv:2211.14966, 2022
152022
Understanding Adversarial Robustness Against On-manifold Adversarial Examples
J Xiao, L Yang, Y Fan, J Wang, ZQ Luo
arXiv preprint arXiv:2210.00430, 2022
102022
PAC-bayesian spectrally-normalized bounds for adversarially robust generalization
J Xiao, R Sun, ZQ Luo
Advances in Neural Information Processing Systems 36, 36305-36323, 2023
8*2023
Improving Adversarial Training for Multiple Perturbations through the Lens of Uniform Stability
J Xiao, Z Qin, Y Fan, B Wu, J Wang, ZQ Luo
ICML 2023 Workshop AdvML-Frontiers, 2023
7*2023
Smoothed-SGDmax: A Stability-Inspired Algorithm to Improve Adversarial Generalization
J Xiao, J Zhang, ZQ Luo, AE Ozdaglar
NeurIPS 2022 ML Safety Workshop, 2022
6*2022
On the Algorithmic Bias of Aligning Large Language Models with RLHF: Preference Collapse and Matching Regularization
J Xiao, Z Li, X Xie, E Getzen, C Fang, Q Long, WJ Su
arXiv preprint arXiv:2405.16455, 2024
42024
Entropic Distribution Matching in Supervised Fine-tuning of LLMs: Less Overfitting and Better Diversity
Z Li, C Chen, T Xu, Z Qin, J Xiao, R Sun, ZQ Luo
arXiv preprint arXiv:2408.16673, 2024
2024
Fine-Tuning Linear Layers Only Is a Simple yet Effective Way for Task Arithmetic
R Jin, B Hou, J Xiao, W Su, L Shen
arXiv preprint arXiv:2407.07089, 2024
2024
Bridging the Gap: Rademacher Complexity in Robust and Standard Generalization
J Xiao, R Sun, Q Long, WJ Su
37th Annual Conference on Learning Theory (COLT 2024) 247, 5074-5075, 2024
2024
Uniformly Stable Algorithms for Adversarial Training and Beyond
J Xiao, J Zhang, ZQ Luo, A Ozdaglar
41th International Conference on Machine Learning (ICML 2024), 2024
2024
El sistema no puede realizar la operación en estos momentos. Inténtalo de nuevo más tarde.
Artículos 1–11