Julia Hockenmaier

Scaling Evaluation-time Compute with Reasoning Models as Process Evaluators

Analyzing Multilingualism in Large Language Models with Sparse Autoencoders

Evaluating and Designing Sparse Autoencoders by Approximating Quasi-Orthogonality

Measuring the Reliability of Causal Probing Methods: Tradeoffs, Limitations, and the Plight of Nullifying Interventions