Choosing the Right Representations: Task-Aware Fusion for Multimodal Materials Language Modelswith Huu Binh Ta, Alex Morehead, Ferdinando Fioretto, Sandeep Madireddy1 versionW_ka87zmtrv1 · current