Yisen Wang

6

Papers

40

Total Citations

Papers (6)

Do Generated Data Always Help Contrastive Learning?

Beyond Interpretability: The Gains of Feature Monosemanticity on Model Robustness

An Augmentation-Aware Theory for Self-Supervised Contrastive Learning

PID: Prompt-Independent Data Protection Against Latent Diffusion Models

TERD: A Unified Framework for Safeguarding Diffusion Models Against Backdoors

Look Ahead or Look Around? A Theoretical Comparison Between Autoregressive and Masked Pretraining