harvard-university

This paper introduces a new embedding model designed to capture preferential similarity rather than just semantic similarity, improving preference prediction for collective decision-making systems.