Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens

LRMRe-assign community