Alignment between human and AI minds

Schematic of two agents (Alice and Bob) seeing the same set of images, forming internal representations, and computing pairwise similarities; the question is whether their representations align. — Two minds looking at the same world. We measure when human and AI representations align, and what that alignment buys us.

The lab’s theoretical anchor. We study how minds, biological and artificial, come to share structure across representations, concepts, values, and perception, and we use that structure as a tool to make AI systems more compatible with people.

Representative work. Alignment with human representations supports robust few-shot learning (NeurIPS 2023 Spotlight) shows a U-shaped relationship between human-model representational alignment and downstream task performance. Getting aligned on representational alignment (TMLR 2025) lays out a community-wide research agenda for the field. On the informativeness of supervision signals (UAI 2023 Spotlight) develops the information-theoretic backbone connecting soft-label supervision to alignment.