1 paper
Qinwu Xu, Xiaofu Ma, Yifan Jiang
We study whether multimodal large language models (MLLMs) can leverage crystallographic plane indices (Miller indices) as a structured latent representation for reasoning about fra…