PI at the Weizmann Institute of Science, also with the MIT-IBM Watson AI Lab and IBM Research. PhD from the Hebrew University of Jerusalem.
Works on natural language processing: self-improving AI, how people and AI interact, evaluating language models and making benchmarks reliable, model merging, efficient pretraining, and what language acquisition in humans can teach us about training models.
Instructor of Active AI Reading. Other places to find the work: Hugging Face, OpenReview, ORCID.