DUNE: Fixing Diffusion Models Without Extra Training
DUNE refines diffusion models by catching early-stage noise issues without re-training. It's a breakthrough for fidelity and hallucinations.
Read the full articleYou might also wanna read
Training-Free Single-Image Diffusion Model Achieves Fast, High-Quality Generation
We consider the problem of generating images whose internal structure -- defined by the distribution of patches across multiple scales -- ma
arxiv.org·1mo ago
Three training-time interventions improve diffusion-based speculative decoding by 21-76%
Large language models (LLMs) achieve remarkable performance across a wide range of tasks, but their autoregressive decoding process incurs s
arxiv.org·21d agoTowards Demystifying the Creativity of Diffusion Models
Algorithms & Theory.
Flow Maps: Accelerating Diffusion Model Sampling by Learning the Integral Directly
A deep dive on flow maps.
Sander Dieleman·21d agoFlow Maps: Accelerating Diffusion Model Sampling by Learning the Integral Directly
A deep dive on flow maps.
Sander Dieleman·21d ago
OpenAI Research Explains Why Language Models Hallucinate and How to Improve Reliability
OpenAI’s new research explains why language models hallucinate. The findings show how improved evaluations can enhance AI reliability, hones
openai.com·10mo ago
Towards demystifying the creativity of diffusion models
Score smoothing facilitates manifold recovery In the real world, complex data like high-resolution images live in high-dimensional pixel spa
Comments
Sign in to join the conversation.
No comments yet. Be the first.