Buchumschlag

On the Clean Generalization and Robust Overfitting in Adversarial Training from Two Theoretical Views: Representation Complexity and Training Dynamics

Fuente: arXiv

Gespeichert in:

Bibliographische Detailangaben
Hauptverfasser:	Li, Binghui, Li, Yuanzhi
Format:	Preprint
Veröffentlicht:	2023
Schlagworte:	Machine Learning
Online-Zugang:	Acceder al recurso
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Ähnliche Einträge

Adversarial Training Can Provably Improve Robustness: Theoretical Analysis of Feature Learning Process Under Structured Data
von: Li, Binghui, et al.
Veröffentlicht: (2024)

On the Impact of Hard Adversarial Instances on Overfitting in Adversarial Training
von: Liu, Chen, et al.
Veröffentlicht: (2021)

Benign Overfitting in Adversarial Training for Vision Transformers
von: Zhang, Jiaming, et al.
Veröffentlicht: (2026)

Revisiting the Relationship between Adversarial and Clean Training: Why Clean Training Can Make Adversarial Training Better
von: Zhou, MingWei, et al.
Veröffentlicht: (2025)

Understanding Generalization in Transformers: Error Bounds and Training Dynamics Under Benign and Harmful Overfitting
von: Zhang, Yingying, et al.
Veröffentlicht: (2025)

Better Representations via Adversarial Training in Pre-Training: A Theoretical Perspective
von: Xing, Yue, et al.
Veröffentlicht: (2024)

Trained Transformer Classifiers Generalize and Exhibit Benign Overfitting In-Context
von: Frei, Spencer, et al.
Veröffentlicht: (2024)

Measure-Theoretic Anti-Causal Representation Learning
von: Behnam, Arman, et al.
Veröffentlicht: (2025)

Unveil Benign Overfitting for Transformer in Vision: Training Dynamics, Convergence, and Generalization
von: Jiang, Jiarui, et al.
Veröffentlicht: (2024)

Benign Overfitting for Regression with Trained Two-Layer ReLU Networks
von: Park, Junhyung, et al.
Veröffentlicht: (2024)

The Surprising Harmfulness of Benign Overfitting for Adversarial Robustness
von: Hao, Yifan, et al.
Veröffentlicht: (2024)

Muon in Associative Memory Learning: Training Dynamics and Scaling Laws
von: Li, Binghui, et al.
Veröffentlicht: (2026)

Improving Clean Accuracy via a Tangent-Space Perspective on Adversarial Training
von: Yi, Bongsoo, et al.
Veröffentlicht: (2024)

Theoretical Analysis of Robust Overfitting for Wide DNNs: An NTK Approach
von: Fu, Shaopeng, et al.
Veröffentlicht: (2023)

Unveiling the Backdoor Mechanism Hidden Behind Catastrophic Overfitting in Fast Adversarial Training
von: Zhao, Mengnan, et al.
Veröffentlicht: (2026)

Inf2Guard: An Information-Theoretic Framework for Learning Privacy-Preserving Representations against Inference Attacks
von: Noorbakhsh, Sayedeh Leila, et al.
Veröffentlicht: (2024)

Preventing Catastrophic Overfitting in Fast Adversarial Training: A Bi-level Optimization Perspective
von: Wang, Zhaoxin, et al.
Veröffentlicht: (2024)

Understanding Robust Overfitting from the Feature Generalization Perspective
von: Yu, Chaojian, et al.
Veröffentlicht: (2023)

Learning Robust and Privacy-Preserving Representations via Information Theory
von: Zhang, Binghui, et al.
Veröffentlicht: (2024)

MEAT: Median-Ensemble Adversarial Training for Improving Robustness and Generalization
von: Hu, Zhaozhe, et al.
Veröffentlicht: (2024)

On the Robustness of Adversarial Training Against Uncertainty Attacks
von: Ledda, Emanuele, et al.
Veröffentlicht: (2024)

Dynamical Decoupling of Generalization and Overfitting in Large Two-Layer Networks
von: Montanari, Andrea, et al.
Veröffentlicht: (2025)

Ignition Phase : Standard Training for Fast Adversarial Robustness
von: Yu-Hang, Wang, et al.
Veröffentlicht: (2025)

Dynamic Label Adversarial Training for Deep Learning Robustness Against Adversarial Attacks
von: Liu, Zhenyu, et al.
Veröffentlicht: (2024)

Learning from Peers: Collaborative Ensemble Adversarial Training
von: Dengjin, Li, et al.
Veröffentlicht: (2025)

Latent Adversarial Training Improves the Representation of Refusal
von: Abbas, Alexandra, et al.
Veröffentlicht: (2025)

TCRL: Temporal-Coupled Adversarial Training for Robust Constrained Reinforcement Learning in Worst-Case Scenarios
von: Xu, Wentao, et al.
Veröffentlicht: (2026)

Stability and Generalization in Free Adversarial Training
von: Cheng, Xiwei, et al.
Veröffentlicht: (2024)

Catastrophic Overfitting, Entropy Gap and Participation Ratio: A Noiseless $l^p$ Norm Solution for Fast Adversarial Training
von: Mehouachi, Fares B., et al.
Veröffentlicht: (2025)

Adapting to Evolving Adversaries with Regularized Continual Robust Training
von: Dai, Sihui, et al.
Veröffentlicht: (2025)

Training Dynamics Impact Post-Training Quantization Robustness
von: Catalan-Tatjer, Albert, et al.
Veröffentlicht: (2025)

A Theoretical Analysis of Mamba's Training Dynamics: Filtering Relevant Features for Generalization in State Space Models
von: Shandirasegaran, Mugunthan, et al.
Veröffentlicht: (2026)

Information Theoretic Adversarial Training of Large Language Models
von: Zhang, Yiwei, et al.
Veröffentlicht: (2026)

Vulnerability-Aware Robust Multimodal Adversarial Training
von: Zhang, Junrui, et al.
Veröffentlicht: (2025)

When Muon Optimizer Meets Adversarial Training: A Theoretical and Empirical Study
von: Yan, Jun, et al.
Veröffentlicht: (2026)

An Agnostic View on the Cost of Overfitting in (Kernel) Ridge Regression
von: Zhou, Lijia, et al.
Veröffentlicht: (2023)

SemRoDe: Macro Adversarial Training to Learn Representations That are Robust to Word-Level Attacks
von: Formento, Brian, et al.
Veröffentlicht: (2024)

On the Noise Stability and Robustness of Adversarially Trained Networks on NVM Crossbars
von: Tao, Chun, et al.
Veröffentlicht: (2021)

Enhancing Evolving Domain Generalization through Dynamic Latent Representations
von: Xie, Binghui, et al.
Veröffentlicht: (2024)

Representation Learning of Multivariate Time Series using Attention and Adversarial Training
von: Scharwächter, Leon, et al.
Veröffentlicht: (2024)