Amit Dhurandhar

CoFrGeNet: Continued Fraction Architectures for Language Generation

Scalable Circuit Learning for Interpreting Large Language Models

EVALUATING LLM SAFETY ACROSS CHILD DEVELOPMENT STAGES: A SIMULATED AGENT APPROACH

Programming Refusal with Conditional Activation Steering

Final-Model-Only Data Attribution with a Unifying View of Gradient-Based Methods

Large Language Model Confidence Estimation via Black-Box Access

Identifying Sub-networks in Neural Networks via Functionally Similar Representations

CELL your Model: Contrastive Explanations for Large Language Models