AES · Research community
Boston Working Group
A recurring research discussion on AI agent evaluation.
About the working group
A small research forum for agent evaluation
The Boston Working Group brings together researchers studying and building AI agent evaluations across universities, industry, and related fields.
Sessions begin with brief introductions to current work, followed by an open discussion of research problems, gaps in methods and infrastructure, and possible collaborations. Attendance is limited to 15 people with prior experience in agent evaluation and requires host approval.
Recent discussions