Yifei Wang

8

Papers

79

Total Citations

Papers (8)

Do Generated Data Always Help Contrastive Learning?

On the Emergence of Position Bias in Transformers

Uni-Instruct: One-step Diffusion Model through Unified Diffusion Divergence Instruction

Beyond Interpretability: The Gains of Feature Monosemanticity on Model Robustness

Semi-IIN: Semi-Supervised Intra-Inter Modal Interaction Learning Network for Multimodal Sentiment Analysis

OODRobustBench: a Benchmark and Large-Scale Analysis of Adversarial Robustness under Distribution Shift

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

On the Duality Between Sharpness-Aware Minimization and Adversarial Training