Performance Plateaus in Inference-Time Scaling for Text-to-Image Diffusion Without External Models

DiffMRe-assign community