Search

Walter Gerych

Complementing Self-Consistency with Cross-Model Disagreement for Uncertainty Quantification
When Style Breaks Safety: Defending LLMs Against Superficial Style Alignment
WRING Out The Bias: A Rotation-Based Alternative To Projection Debiasing
An Investigation of Memorization Risk in Healthcare Foundation Models
Opportunistic screening of type 2 diabetes with deep metric learning using electronic health records
Test-Time Debiasing of Vision-Language Embeddings