Berkeley correction training cuts diffusion exposure bias, high-quality samples without guidance
UCBerkeley · hf · 2026-10-21
A paper from UC Berkeley and Reve trains a correction module that refines diffusion model predictions, reducing exposure bias between training and sampling.
- Achieves high-quality class-conditional and unconditional ImageNet samples without guidance (no CFG)
- The correction module can be distilled away, eliminating inference overhead
- Targets the common weakness of poor unguided sample quality
More from Multimodal
- EvaOmni: an evidence-grounded benchmark for audio-visual reasoning in omnimodal LLMs — VisionXLab · 2026-10-21
- WinARDet: compressed window autoregression stops box repetition loops in dense grounding — NankaiUniversity · 2026-10-21
- Qwen-Image 2.1 edit keeps producing plastic skin, negative prompts don't help — NorthAppropriate4458 · 2026-10-21
- AI-generated illustration visualizes Paz Soldán's sci-fi story 'Artificial' and its Ship of Theseus theme — dasein47 · 2026-10-21
- AI-generated Halloween video of a woman walking through skeleton crowds goes viral — Stephane_AI · 2026-10-21
- AI-generated fantasy short Lost In Woods stuns with cinematic forest fairy scene — Some-Dark-5802 · 2026-10-21