Resources

Recommended Reading

Foundation

• A Connection Between Score Matching and Denoising Autoencoders

• Score-Based Generative Modeling through Stochastic Differential Equations

• Elucidating the Design Space of Diffusion-Based Generative Models

• The Principles of Diffusion Models

• Classifier-Free Diffusion Guidance

• Guiding a Diffusion Model with a Bad Version of Itself

• Flow Matching Guide and Code

• Denoising Diffusion Probabilistic Models

• Variational Diffusion Models

• Diffusion Models and Gaussian Flow Matching: Two Sides of the Same Coin

• Diffusion Models Beat GANs on Image Synthesis

• Stochastic Interpolants: A Unifying Framework for Flows and Diffusions

• Generalization in diffusion models arises from geometry-adaptive harmonic representations

• Memorization and Regularization in Generative Diffusion Models

Scalable Pretraining

• Analyzing and Improving the Training Dynamics of Diffusion Models

• Scalable Diffusion Models with Transformers

• All are Worth Words: A ViT Backbone for Diffusion Models

• Fast Training of Diffusion Models with Masked Transformers

• SiT: Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers

• Scaling Rectified Flow Transformers for High-Resolution Image Synthesis

• Back to Basics: Let Denoising Generative Models Denoise

• Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Models

• Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

• Self-Supervised Flow Matching for Scalable Multi-Modal Synthesis

Fast Sampling

• DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps

• Progressive Distillation for Fast Sampling of Diffusion Models

• Consistency Models

• Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion

• How to build a consistency model: Learning flow maps via self-distillation

• Improved Distribution Matching Distillation for Fast Image Synthesis

• Mean Flows for One-step Generative Modeling

Post-Training

• Diffusion Model Alignment Using Direct Preference Optimization

• Flow-GRPO: Training Flow Matching Models via Online RL

• Diffusion Policy Policy Optimization

• DiffusionNFT: Online Diffusion Reinforcement with Forward Process

• Rethinking the Design Space of Reinforcement Learning for Diffusion Models: On the Importance of Likelihood Estimation Beyond Loss Design

Test-time Scaling & Inverse Problems

• A General Framework for Inference-time Scaling and Steering of Diffusion Models

• Inference-Time Text-to-Video Alignment with Diffusion Latent Beam Search

• Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps

• Diffusion Posterior Sampling for General Noisy Inverse Problems

• Zero-Shot Image Restoration Using Denoising Diffusion Null-Space Model

• A Variational Perspective on Solving Inverse Problems with Diffusion Models

• Improving Diffusion Inverse Problem Solving with Decoupled Noise Annealing

• Principled Probabilistic Imaging using Diffusion Models as Plug-and-Play Priors

• InverseBench: Benchmarking Plug-and-Play Diffusion Priors for Inverse Problems in Physical Sciences

• Blade: A Derivative-free Bayesian Inversion Method using Diffusion Priors

• STeP: A Framework for Solving Scientific Video Inverse Problems with Spatiotemporal Diffusion Priors

Applications in Vision

• FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space

• FLUX.2: Analyzing and Enhancing the Latent Space of FLUX – Representation Comparison

• LTX-2: Efficient Joint Audio-Visual Foundation Model

• Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer

• Wan: Open and Advanced Large-Scale Video Generative Models

• DreamFusion: Text-to-3D using 2D Diffusion

• SAM 3D: 3Dfy Anything in Images

• Native and Compact Structured Latents for 3D Generation

• DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation

Physical AI

• Diffusion Policy: Visuomotor Policy Learning via Action Diffusion

• π0: A Vision-Language-Action Flow Model for General Robot Control

• GR00T N1: An Open Foundation Model for Generalist Humanoid Robots

• Cosmos World Foundation Model Platform for Physical AI

• Genie 2: A large-scale foundation world model

Related Courses

• Stanford CS236 Deep Generative Models

• CMU 10-799 Diffusion & Flow Matching

• CMU 18-789 Deep Generative Modeling