Picture for Haichang Gao

Haichang Gao

Mining Glitch Tokens in Large Language Models via Gradient-based Discrete Optimization

Add code
Oct 19, 2024
Viaarxiv icon

The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models

Add code
Jul 25, 2024
Figure 1 for The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models
Figure 2 for The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models
Figure 3 for The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models
Figure 4 for The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models
Viaarxiv icon

SoK: Acoustic Side Channels

Add code
Aug 06, 2023
Viaarxiv icon

AdvFunMatch: When Consistent Teaching Meets Adversarial Robustness

Add code
May 25, 2023
Viaarxiv icon

Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training

Add code
Aug 26, 2022
Figure 1 for Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
Figure 2 for Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
Figure 3 for Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
Figure 4 for Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
Viaarxiv icon

Alleviating Robust Overfitting of Adversarial Training With Consistency Regularization

Add code
May 24, 2022
Figure 1 for Alleviating Robust Overfitting of Adversarial Training With Consistency Regularization
Figure 2 for Alleviating Robust Overfitting of Adversarial Training With Consistency Regularization
Figure 3 for Alleviating Robust Overfitting of Adversarial Training With Consistency Regularization
Figure 4 for Alleviating Robust Overfitting of Adversarial Training With Consistency Regularization
Viaarxiv icon