Satoshi Hayakawa

SONA: Learning Conditional, Unconditional, and Matching-Aware Discriminator

Concept-TRAK: Understanding how diffusion models learn concepts through concept attribution

Theoretical refinement of CLIP by utilizing linear structure of optimal similarity

Demystifying MaskGIT Sampler and Beyond: Adaptive Order Selection in Masked Diffusion

MCA: Modality Composition Awareness for Robust Composed Multimodal Retrieval

Jump Your Steps: Optimizing Sampling Schedule of Discrete Diffusion Models

Distillation of Discrete Diffusion through Dimensional Correlations