AI RESEARCH
Rethinking Token Reduction for Diffusion Models via Output-Similarity-Awareness
arXiv CS.CV
•
ArXi:2605.22011v1 Announce Type: new Diffusion Transformers (DiTs) achieve superior image generation quality but suffer from quadratic computational complexity relative to token count. While various token reduction (TR) methods have been proposed to mitigate this cost, they overlook the primary objective of generative models: minimizing recovery error, which requires reflecting output token similarity. They rely solely on input token similarity inherited from reduction-only ViT paradigms, leading to a fundamental misalignment with this objective.