Learning Representations to Guide Pixel Diffusion Transformers

arXiv:2610.12978 · cs.CV · Submitted 2026-10-09 · Read on arXiv

cs.CV

Submitted: 2026-10-09

Updated: 2026-10-09

Comments: 26 pages. Code: https://github.com/Pan-Zihao/Representation-Guidance Project page: https://pan-zihao.github.io/Representation-Guidance

Code: https://github.com/Pan-Zihao/Representation-Guidance

Project page: https://pan-zihao.github.io/Representation-Guidance

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Transformer — 1× in this paper · explained in 2 episodes

Sources

• BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset
• ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation
• PixelFlow: Pixel-Space Generative Models with Flow
• Scaling Instruction-Finetuned Language Models
• PixelU: A U-Shaped Transformer for Efficient End-to-End Pixel Diffusion
• Classifier-Free Diffusion Guidance
• ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
• Scalable Adaptive Computation for Iterative Generation
• PixelGen: Improving Pixel Diffusion with Perceptual Supervision
• DINOv2: Learning Robust Visual Features without Supervision
• Vision Transformers Need More Than Registers
• Representation Alignment for Just Image Transformers is not Easier than You Think
• DINOv3
• Improved Baselines with Representation Autoencoders
• Diffusion Models without Classifier-free Guidance
• JetFormer: An Autoregressive Generative Model of Raw Images and Text
• DDT: Decoupled Diffusion Transformer
• Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think
• What Matters for Diffusion-Friendly Latent Manifold? Prior-Aligned Autoencoders for Latent Diffusion
• Training-Free Representation Guidance for Diffusion Models with a Representation Alignment Projector

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models