DeCo: Faster, sharper pixel diffusion by splitting frequencies

Kari Jaaskelainen

25 Nov 2025 — 1 min read

Meet DeCo, a new way to generate images directly in pixels - faster and sharper.

Instead of one big model juggling everything, DeCo splits the job by "frequencies": the Diffusion Transformer learns low-frequency stuff (overall shapes and colors), while a lightweight pixel decoder paints high-frequency details (edges and textures). Think bass vs treble, now tuned separately.

Better focus: A frequency-aware flow-matching loss emphasizes visually important signals and downplays noise.
Results: ImageNet FID 1.62 (256x256) and 2.22 (512x512) - closing the gap with latent diffusion. Their text-to-image model hits 0.86 on GenEval.
Open code: https://github.com/Zehong-Ma/DeCo

Takeaway: decoupling low-level details from global structure makes end-to-end pixel diffusion more efficient without sacrificing realism.

Paper: https://arxiv.org/abs/2511.19365v1

Paper: https://arxiv.org/abs/2511.19365v1

Register: https://www.AiFeta.com

#AI #Diffusion #ImageGeneration #ComputerVision #DeepLearning #OpenSource #Research #ImageNet

Meet GRACE: a moral governor for safer, more transparent AI

AI agents are getting powerful—so how do we make sure they do the right thing, not just the effective thing? Meet GRACE, a reason-based moral governor that keeps AI behavior aligned with human norms by separating moral reasoning from goal-driven decision-making. * Moral Module: uses deontic logic and explicit reasons

Why some fine-tuned LLMs miss phishing—and how to fix it

Not all fine-tuned LLMs spot phishing equally. A new study tests Llama 3.1 8B, Gemma 2 9B, and Mistral on high-stakes phishing detection—and uses SHAP and mechanistic interpretability to reveal why models do (or don’t) generalize. * Architecture × data diversity matters: Gemma 2 9B hits state-of-the-art performance (F1

AI that explains itself—by following the science

AI that explains itself—by following the science Black-box AI can be powerful, but it often can’t tell us why it made a decision. Concept Bottleneck Models (CBMs) try to fix that by predicting human-understandable concepts first, then the final answer. The catch: standard CBMs ignore domain-specific cause-and-effect and

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A

TL;DR: Finding reliable answers in genomics is hard. GenomAgent turns one big AI into a coordinating team of specialists—and beats the current leader by 12% on a key benchmark. Genomic facts live across many databases. Standard chatbots struggle because they can’t flexibly query those sources. GeneGPT added

Read more

Meet GRACE: a moral governor for safer, more transparent AI

Why some fine-tuned LLMs miss phishing—and how to fix it

AI that explains itself—by following the science

Meet GenomAgent: A Team of AI Specialists for Smarter Genomics Q&A