Picture for Carmela Troncoso

Carmela Troncoso

EPFL

Attack-Aware Noise Calibration for Differential Privacy

Add code
Jul 02, 2024
Viaarxiv icon

SINBAD: Saliency-informed detection of breakage caused by ad blocking

Add code
May 08, 2024
Figure 1 for SINBAD: Saliency-informed detection of breakage caused by ad blocking
Figure 2 for SINBAD: Saliency-informed detection of breakage caused by ad blocking
Figure 3 for SINBAD: Saliency-informed detection of breakage caused by ad blocking
Figure 4 for SINBAD: Saliency-informed detection of breakage caused by ad blocking
Viaarxiv icon

Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks

Add code
Mar 06, 2024
Figure 1 for Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks
Figure 2 for Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks
Figure 3 for Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks
Figure 4 for Neural Exec: Learning (and Learning from) Execution Triggers for Prompt Injection Attacks
Viaarxiv icon

On the Conflict of Robustness and Learning in Collaborative Machine Learning

Add code
Feb 21, 2024
Viaarxiv icon

The Fundamental Limits of Least-Privilege Learning

Add code
Feb 19, 2024
Figure 1 for The Fundamental Limits of Least-Privilege Learning
Figure 2 for The Fundamental Limits of Least-Privilege Learning
Figure 3 for The Fundamental Limits of Least-Privilege Learning
Figure 4 for The Fundamental Limits of Least-Privilege Learning
Viaarxiv icon

Transferable Adversarial Robustness for Categorical Data via Universal Robust Embeddings

Add code
Jun 06, 2023
Viaarxiv icon

Can Decentralized Learning be more robust than Federated Learning?

Add code
Mar 07, 2023
Viaarxiv icon

Arbitrary Decisions are a Hidden Cost of Differentially-Private Training

Add code
Feb 28, 2023
Figure 1 for Arbitrary Decisions are a Hidden Cost of Differentially-Private Training
Figure 2 for Arbitrary Decisions are a Hidden Cost of Differentially-Private Training
Figure 3 for Arbitrary Decisions are a Hidden Cost of Differentially-Private Training
Figure 4 for Arbitrary Decisions are a Hidden Cost of Differentially-Private Training
Viaarxiv icon

Universal Neural-Cracking-Machines: Self-Configurable Password Models from Auxiliary Data

Add code
Jan 18, 2023
Viaarxiv icon

Adversarial Robustness for Tabular Data through Cost and Utility Awareness

Add code
Aug 27, 2022
Figure 1 for Adversarial Robustness for Tabular Data through Cost and Utility Awareness
Figure 2 for Adversarial Robustness for Tabular Data through Cost and Utility Awareness
Figure 3 for Adversarial Robustness for Tabular Data through Cost and Utility Awareness
Figure 4 for Adversarial Robustness for Tabular Data through Cost and Utility Awareness
Viaarxiv icon