Agent Evaluation Science Boston Working Group

​The Boston Agent Evaluation Working Group brings together researchers, mainly PhD students, across different universities in the broader Boston/Cambridge area who are actively building or studying AI agent evaluations.

​Participation in this working group is limited to people with prior experience in agent evaluation. Everyone will briefly introduce what they are working on, followed by an open discussion of important problems in agent evaluation, gaps in current methods and infrastructure, and directions that may be worth pursuing together.

​The goal is to create a small, recurring research community where people can exchange work in progress, develop collaborations, and connect with the broader Northeast agent evaluation community.

​To keep the discussion focused and substantive, the group is limited to 10 participants. We will review the information submitted by all applicants. If we are unable to accommodate you in this session, we will keep your information on file and prioritize you for a future working group meeting.