Shauli Ravfogel

Beyond Single Embeddings: Capturing Diverse Targets with Multi-Query Retrieval

Emergence of Linear Truth Encodings in Language Models

Preserving Task-Relevant Information Under Linear Concept Removal

Gumbel Counterfactual Generation From Language Models

Intrinsic Evaluation of Unlearning Using Parametric Knowledge Traces

GRADE: Quantifying Sample Diversity in Text-to-Image Models