Picture for Guillaume Leclerc

Guillaume Leclerc

Rethinking Backdoor Attacks

Add code
Jul 19, 2023
Figure 1 for Rethinking Backdoor Attacks
Figure 2 for Rethinking Backdoor Attacks
Figure 3 for Rethinking Backdoor Attacks
Figure 4 for Rethinking Backdoor Attacks
Viaarxiv icon

FFCV: Accelerating Training by Removing Data Bottlenecks

Jun 21, 2023
Figure 1 for FFCV: Accelerating Training by Removing Data Bottlenecks
Figure 2 for FFCV: Accelerating Training by Removing Data Bottlenecks
Figure 3 for FFCV: Accelerating Training by Removing Data Bottlenecks
Figure 4 for FFCV: Accelerating Training by Removing Data Bottlenecks
Viaarxiv icon

TRAK: Attributing Model Behavior at Scale

Add code
Apr 03, 2023
Figure 1 for TRAK: Attributing Model Behavior at Scale
Figure 2 for TRAK: Attributing Model Behavior at Scale
Figure 3 for TRAK: Attributing Model Behavior at Scale
Figure 4 for TRAK: Attributing Model Behavior at Scale
Viaarxiv icon

Raising the Cost of Malicious AI-Powered Image Editing

Add code
Feb 13, 2023
Figure 1 for Raising the Cost of Malicious AI-Powered Image Editing
Figure 2 for Raising the Cost of Malicious AI-Powered Image Editing
Figure 3 for Raising the Cost of Malicious AI-Powered Image Editing
Figure 4 for Raising the Cost of Malicious AI-Powered Image Editing
Viaarxiv icon

Adversarially trained neural representations may already be as robust as corresponding biological neural representations

Add code
Jun 19, 2022
Figure 1 for Adversarially trained neural representations may already be as robust as corresponding biological neural representations
Figure 2 for Adversarially trained neural representations may already be as robust as corresponding biological neural representations
Figure 3 for Adversarially trained neural representations may already be as robust as corresponding biological neural representations
Figure 4 for Adversarially trained neural representations may already be as robust as corresponding biological neural representations
Viaarxiv icon

Datamodels: Predicting Predictions from Training Data

Add code
Feb 01, 2022
Figure 1 for Datamodels: Predicting Predictions from Training Data
Figure 2 for Datamodels: Predicting Predictions from Training Data
Figure 3 for Datamodels: Predicting Predictions from Training Data
Figure 4 for Datamodels: Predicting Predictions from Training Data
Viaarxiv icon

3DB: A Framework for Debugging Computer Vision Models

Add code
Jun 07, 2021
Figure 1 for 3DB: A Framework for Debugging Computer Vision Models
Figure 2 for 3DB: A Framework for Debugging Computer Vision Models
Figure 3 for 3DB: A Framework for Debugging Computer Vision Models
Figure 4 for 3DB: A Framework for Debugging Computer Vision Models
Viaarxiv icon

Revisiting Ensembles in an Adversarial Context: Improving Natural Accuracy

Feb 26, 2020
Figure 1 for Revisiting Ensembles in an Adversarial Context: Improving Natural Accuracy
Figure 2 for Revisiting Ensembles in an Adversarial Context: Improving Natural Accuracy
Viaarxiv icon

The Two Regimes of Deep Network Training

Feb 24, 2020
Figure 1 for The Two Regimes of Deep Network Training
Figure 2 for The Two Regimes of Deep Network Training
Figure 3 for The Two Regimes of Deep Network Training
Figure 4 for The Two Regimes of Deep Network Training
Viaarxiv icon

Smallify: Learning Network Size while Training

Jun 10, 2018
Figure 1 for Smallify: Learning Network Size while Training
Figure 2 for Smallify: Learning Network Size while Training
Viaarxiv icon