Sean O'Brien
Principal Investigator
Sean O'Brien is a UCSD PhD student in the McAuley Lab researching how to help language models reason better. He started at UC Berkeley in 2018, studying applied math, cognitive science, electrical engineering, and computer science.
During his time at Meta AI, Sean co-authored Shepherd, a language model specifically tuned to critique responses and suggest refinements. Even though Shepherd is small (7B parameters), its critiques are either equivalent or preferred to those from established models including ChatGPT. In human evaluation, Shepherd strictly outperforms other models and closely ties with ChatGPT.
Sean also co-authored "Contrastive Decoding Improves Reasoning in Large Language Models" and "PathFinder: Guided Search over Multi-Step Reasoning Paths."
At Berkeley, he taught 7 GSI positions covering programming, discrete mathematics, and machine learning, training the next generation of computer scientists. He is passionate about both advancing AI research and teaching students to do the same.
Begin Your Journey
The application takes 10 minutes and is reviewed on a rolling basis. We look for strong technical signal—projects, coursework, or competition results—and a genuine curiosity to do real research.
If admitted, you will join a structured pipeline with direct mentorship to take your work from ideation to top conference submission at venues like NeurIPS, ACL, and EMNLP.
