Exploring Human-AI Conceptual Alignment through the Prism of Chess
Abstract
Using chess and Chess960, we investigate how strategic concepts appear across model layers and how stronger play can diverge from human conceptual understanding.
Type
Publication
NeurIPS 2025 Creative AI Track
Ravid Shwartz-Ziv
AI Researcher
AI researcher at Meta MSL with a background in information theory and computational neuroscience, working on world models, memory, and compression. Former Assistant Professor and Faculty Fellow at NYU, collaborating on research across academia and industry.