Picture for Wassim Bouaziz

Wassim Bouaziz

A Visual Case Study of the Training Dynamics in Neural Networks

Add code
Oct 31, 2024
Viaarxiv icon

Inverting Gradient Attacks Naturally Makes Data Poisons: An Availability Attack on Neural Networks

Add code
Oct 28, 2024
Figure 1 for Inverting Gradient Attacks Naturally Makes Data Poisons: An Availability Attack on Neural Networks
Figure 2 for Inverting Gradient Attacks Naturally Makes Data Poisons: An Availability Attack on Neural Networks
Figure 3 for Inverting Gradient Attacks Naturally Makes Data Poisons: An Availability Attack on Neural Networks
Figure 4 for Inverting Gradient Attacks Naturally Makes Data Poisons: An Availability Attack on Neural Networks
Viaarxiv icon

Data Taggants: Dataset Ownership Verification via Harmless Targeted Data Poisoning

Add code
Oct 09, 2024
Viaarxiv icon

Iteration Head: A Mechanistic Study of Chain-of-Thought

Add code
Jun 04, 2024
Viaarxiv icon