Yifeng Xu

openvla. OpenVLA: An open-source vision-language-action model for robotic manipulation.

octo. Octo is a transformer-based robot policy trained on a diverse mix of 800k robot trajectories.

Self-Flow. [ICML'26] Code and website for Self-Flow: Self-Supervised Flow Matching for Scalable Multi-Modal Synthesis

awesome-vla-wam. A Curated List of Vision-Language-Action (VLA) and World Action Models (WAM) Research and Beyond

mira. Code for MIRA: Multiplayer Interactive World Models with Representation Autoencoders

minit2i-torch. Official PyTorch re-implementation of MiniT2I.

UniVideo. [ICLR 2026] UniVideo: Unified Understanding, Generation, and Editing for Videos

Echo-Memory. A Simple Baseline for Video World Models with Memory

vggt. [CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer

rcm. rCM & Causal-rCM: Leading and Unified Algorithms/Infrastructures for Bidirectional/Autoregressive Video Diffusion Distillation at Scale

WEM. Python

mochi. The best OSS video generation models, created by Genmo

HunyuanVideo-1.5. HunyuanVideo-1.5: A leading lightweight video generation model

codex. Lightweight coding agent that runs in your terminal

PiD. PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion

kairos. Official code for world model Kairos

HiDream-O1-Image. Python

SkyReels-V3. SkyReels V3: Multimodal Video Generation Model

SkyReels-V2. SkyReels-V2: Infinite-length Film Generative model

LongLive. Long Video Gen Infrastructure

RAEv2. Official Implemenation for RAEv2: Improved Baselines with Representation Autoencoders

minWM. A Minimal and Elegant Framework & Tutorial for Real-Time Interactive World Models

Awesome-Video-World-Models-with-AR-Diffusion. A Curated List of Awesome Video World Models with AR Diffusion: Covering Algorithms, Applications, and Infrastructure, Aimed at Serving as a Comprehensive Resource for Researchers, Practitioners, and Enthusiasts.

CausVid. (CVPR 2025) From Slow Bidirectional to Fast Autoregressive Video Diffusion Models

Causal-Forcing. [ICML 2026] Official codebase for "Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation" & Causal Forcing++

FD-Loss. Python

SenseNova-U1. SenseNova-U series: Native Unified Paradigm with NEO-unify from the First Principles

tuna-2. Official implementation of Tuna-2: Pixel Embeddings Beat Vision Encoders for Unified Understanding and Generation

decord. An efficient video loader for deep learning with smart shuffling that's super easy to digest

patch-forcing. [CVPR 2026] Denoising, Fast and Slow: Difficulty-Aware Adaptive Sampling for Image Generation

Astra. [ICLR 2026] Astra : General Interactive World Model with Autoregressive Denoising"

YUME. The official code of Yume

lingbot-world. Advancing Open-source World Models

trash-cli. Command line interface to the freedesktop.org trashcan.

PixelDiT. [CVPR 2026 Best Paper Finalist] Pixel Diffusion Transformers for Image Generation

GRN. [ECCV 2026] Generative Refinement Networks for Visual Synthesis (Support C2I & T2I & T2V)

flow-maps. Official codebase for the paper "How to build a consistency model: Learning flow maps via self-distillation" (NeurIPS 2025).

Adversarial-Flow-Models. Python

py-meanflow. Pytorch implementation for MeanFlow

MeanFlow. Pytorch implementation of MeanFlow on ImageNet and CIFAR10

stylegan-t. [ICML'23] StyleGAN-T: Unlocking the Power of GANs for Fast Large-Scale Text-to-Image Synthesis

JoyAI-Image. JoyAI-Image is the unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing.

btop. A monitor of resources

edge_eval_python. A python implementation of edge eval

UNITE-tokenization-generation. Single-stage End-to-End Training for Tokenization and Generation

gmmn. Generative moment matching networks

drifting. Python

CubiD. [CVPR2026 Highlight] Cubic Discrete Diffusion: Discrete Visual Generation on High-Dimensional Representation Tokens https://arxiv.org/abs/2603.19232

coulomb_gan. Implementation of Coulomb GANs

drifting-model. Personal PyTorch implementation of "Generative Modeling via Drifting" with Claude

MMD-GAN. MMD-GAN: Towards Deeper Understanding of Moment Matching Network

BitDance. BitDance & UniWeTok: Open-source autoregressive model with binary visual tokens. A research project for building powerful multimodal autoregressive model.

flash-bidirectional-linear-attention. Triton implement of bi-directional (non-causal) linear attention

DiG. [CVPR 2025] DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention

PixelGen. Official repository for “PixelGen: Improving Pixel Diffusion with Perceptual Loss”

BAR. [ICML 2026] code & model for arxiv paper "Autoregressive Image Generation with Masked Bit Modeling"

imeanflow. Official Implementation of iMF https://arxiv.org/abs/2512.02012

pMF. Official Implementation of pMF https://arxiv.org/abs/2601.22158

FastGen. NVIDIA FastGen: Fast Generation from Diffusion Models

iFSQ. iFSQ & LlamaGen-REPA

Scale-RAE. Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders

Self-Transcendence. [ECCV 2026] Official code repository for "Self-transcendence: Is External Feature Guidance Indispensable for Accelerating Diffusion Transformer Training?"

SpeedrunDiT. SR-DiT Speedrunning ImageNet Diffusion

MoGe. [CVPR'25 Oral] MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision

RollingForcing. [ICLR 2026] Official Repo for Rolling Forcing: Autoregressive Long Video Diffusion in Real Time

GameGen-X.

MaskDiT. Code for Fast Training of Diffusion Models with Masked Transformers

Meissonic. [ICLR 2025] Official Implementation of Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

Muon. Muon is an optimizer for hidden layers in neural networks

pixel-perfect-depth. [NeurIPS 2025] Pixel-Perfect Depth

GLM-Image. GLM-Image: Auto-regressive for Dense-knowledge and High-fidelity Image Generation.

SiD. PyTorch code and model checkpoints for Score identity Distillation (SiD) and its adversarial version (SiDA)

REPA-E. [ICCV 2025] Official implementation of the paper: REPA-E: Unlocking VAE for End-to-End Tuning of Latent Diffusion Transformers

nepa. PyTorch implementation of NEPA

FramePack. Lets make video diffusion practical!

NextFlow. NextFlow🚀: Unified Sequential Modeling Activates Multimodal Understanding and Generation

Internal-Guidance. CVPR 2026 (Highlight)-Guiding a Diffusion Transformer with the Internal Dynamics of Itself (IG)

Grounded-SAM-2. Grounded SAM 2: Ground and Track Anything in Videos with Grounding DINO, Florence-2 and SAM 2

py360convert. Python implementation of convertion between equirectangular, cubemap and perspective. (equirect2cube, cube2equirect, equirect2perspec)

ctm. Python

GFT. Python

SoFlow. [ICLR 2026] SoFlow: Solution Flow Models for One-Step Generative Modeling

CoP. The Collapse of Patches

htop. htop - an interactive process viewer

fd. A simple, fast and user-friendly alternative to 'find'

HY-WorldPlay. HY-World 1.5: A Systematic Framework for Interactive World Modeling with Real-Time Latency and Geometric Consistency

FlowEdit. Official implementation of the paper: "FlowEdit: Inversion-Free Text-Based Editing Using Pre-Trained Flow Models"

GRAG-Image-Editing. https://little-misfit.github.io/GRAG-Image-Editing/

torchdiffeq. Differentiable ODE solvers with full GPU support and O(1)-memory backpropagation.

SVG-T2I. [Arxiv 2025] Official PyTorch Implementation of "SVG-T2I: Scaling up Text-to-Image Latent Diffusion Model Without Variational Autoencoder".

JoPano. JoPano: Unified Panorama Generation via Joint Modeling

TwinFlow. [ICLR 2026] Taming large-scale few-step training with self-adversarial flows! 👏🏻

SFD. [CVPR 2026] Semantics Lead the Way: Harmonizing Semantic and Texture Modeling with Asynchronous Latent Diffusion

Awesome-Pixel-Flow.

REG. [NeurIPS 2025 Oral] Representation Entanglement for Generation: Training Diffusion Transformers Is Much Easier Than You Think

progressive_growing_of_gans. Progressive Growing of GANs for Improved Quality, Stability, and Variation

Z-Image. Python

DeCo. [CVPR2026 Highlight] Official repository for “DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation”

flux2. Official inference repo for FLUX.2 models

kandinsky-5. Kandinsky 5.0: A family of diffusion models for Video & Image generation

clash-verge-rev. A modern GUI client based on Tauri, designed to run in Windows, macOS and Linux for tailored proxy experience

improved-aesthetic-predictor. CLIP+MLP Aesthetic Score Predictor

Awesome-Visual-Tokenizer. Awesome Visual Tokenizers/Autoencoders

JiT. PyTorch implementation of JiT https://arxiv.org/abs/2511.13720

DDT. [CVPR 2026] DDT: Decoupled Diffusion Transformer

EPG. [ICLR2026] There is No VAE: End-To-End Pixel-Space Generative Modeling Via Self-Supervised Pre-Training

DanceGRPO. An official implementation of DanceGRPO: Unleashing GRPO on Visual Generation

EditScore. [ICLR 2026] EditScore: Unlocking Online RL for Image Editing via High-Fidelity Reward Modeling

Edit-R1. Edit-R1: Reinforce Image Editing with Diffusion Negative-Aware Finetuning and MLLM Implicit Feedback

pico-banana-400k. Python

GPT-Image-Edit. GPT-IMAGE-EDIT-1.5M: A Million-Scale, GPT-Generated Image Dataset

UniControl. Unified Controllable Visual Generation Model

zsh-z. Jump quickly to directories that you have visited "frecently." A native Zsh port of z.sh with added features.

SVG. [ICLR 2026] Official PyTorch Implementation of "Latent Diffusion Model Without Variational Autoencoder".

HunyuanImage-3.0. HunyuanImage-3.0: A Powerful Native Multimodal Model for Image Generation

FSQ-pytorch. A Pytorch Implementation of Finite Scalar Quantization

ImageReward. [NeurIPS 2023] ImageReward: Learning and Evaluating Human Preferences for Text-to-image Generation

HPSv2. Human Preference Score v2: A Solid Benchmark for Evaluating Human Preferences of Text-to-Image Synthesis

RAE. Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

ReDi. [NeurIPS'25 Spotlight] Boosting Generative Image Modeling via Joint Image-Feature Synthesis

ULMEvalKit. ULMEvalKit: One-Stop Eval ToolKit for Image Generation

termcolor. ANSI color formatting for output in terminal

clean-fid. PyTorch - FID calculation with proper image resizing and quantization steps [CVPR 2022]

IQA-PyTorch. 🔎 🖼️ 🔥PyTorch Toolbox for Image Quality Assessment, including PSNR, SSIM, LPIPS, FID, NIQE, NRQM(Ma), MUSIQ, TOPIQ, NIMA, DBCNN, BRISQUE, PI and more...

rich. Rich is a Python library for rich text and beautiful formatting in the terminal.

Uni-X. [ICLR 2026] Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Models

Lumina-DiMOO. Lumina-DiMOO - An Open-Sourced Multi-Modal Large Diffusion Language Model

ST-AR.

geneval. GenEval: An object-focused framework for evaluating text-to-image alignment

DiffusionNFT. [ICLR 2026 Oral] DiffusionNFT: Online Diffusion Reinforcement with Forward Process

DCGAN-LSGAN-WGAN-GP-DRAGAN-Tensorflow-2. Reimplementation of GANs

SRPO. Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference

bpytop. Linux/OSX/FreeBSD resource monitor

logos. A huge collection of SVG logos

TiM. Transition Models

NExT-GPT. Code and models for ICML 2024 paper, NExT-GPT: Any-to-Any Multimodal Large Language Model

InternVL. [CVPR 2024 Oral] InternVL Family: A Pioneering Open-Source Alternative to GPT-4o. 接近GPT-4o表现的开源多模态对话模型

VLMEvalKit. Open-source evaluation toolkit of large multi-modality models (LMMs), support 220+ LMMs, 80+ benchmarks

lmms-eval. One-for-All Multimodal Evaluation Toolkit Across Text, Image, Video, and Audio Tasks

Awesome-Multimodal-Large-Language-Models. :sparkles::sparkles:Latest Advances on Multimodal Large Language Models

SEED-X. Multimodal Models in Real World

HiDream-I1. Python

HiDream-E1. Python

MagicQuill. [CVPR'25] Official Implementations for Paper - MagicQuill: An Intelligent Interactive Image Editing System

NextStep-1. [🚀 ICLR 2026 Oral] NextStep-1: SOTA Autogressive Image Generation with Continuous Tokens. A research project developed by the StepFun’s Multimodal Intelligence team.

FreeNoise. [ICLR 2024] Code for FreeNoise based on VideoCrafter

fx. Terminal JSON viewer & processor

SRA. [ICLR 2026] Self-Representation Alignment for Diffusion Transformers (SRA)

DispLoss. Python

REPA. [ICLR'25 Oral] Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

platonic-rep. Python

Wan2.2. Wan: Open and Advanced Large-Scale Video Generative Models

dinov3. Reference PyTorch implementation and models for DINOv3

UniPic. Open-source SOTA multi-image editing model

HPSv3. Official implementation of HPSv3: Towards Wide-Spectrum Human Preference Score (ICCV2025)

Tar. [NeurIPS 2025] Vision as a Dialect: Unifying Visual Understanding and Generation via Text-Aligned Representations

Qwen-Image. Qwen-Image is a powerful image generation foundation model capable of complex text rendering and precise image editing.

PixNerd. [ICLR 2026] PixNerd: Pixel Neural Field Diffusion

ml-inrflow. Python

HunyuanWorld-1.0. Generating Immersive, Explorable, and Interactive 3D Worlds from Words or Pixels with Hunyuan3D World Model

SDEdit. PyTorch implementation for SDEdit: Image Synthesis and Editing with Stochastic Differential Equations

GigaTok. [ICCV 2025] Official repo for "GigaTok: Scaling Visual Tokenizers to 3 Billion Parameters for Autoregressive Image Generation"

radial-attention. [NeurIPS 2025] Radial Attention: O(nlogn) Sparse Attention with Energy Decay for Long Video Generation

supervision. We write your reusable computer vision tools. 💜

FUDOKI. [NeurIPS 2025 Spotlight] FUDOKI: Discrete Flow-based Unified Understanding and Generation via Kinetic-Optimal Velocities

ReVQ. Explore how to get a VQ-VAE models efficiently!

Kimi-K2. Kimi K2 is the large language model series developed by Moonshot AI team

aac-datasets. Audio Captioning datasets for PyTorch.

Ovis-U1. An unified model that seamlessly integrates multimodal understanding, text-to-image generation, and image editing within a single powerful framework.

Step1X-Edit. A SOTA open-source image editing model, which aims to provide comparable performance against the closed-source models like GPT-4o and Gemini 2 Flash.

x-flux. Python

dinov2. PyTorch code and models for the DINOv2 self-supervised learning method.

Awesome-Unified-Understanding-and-Generation.

AnyGPT. Code for "AnyGPT: Unified Multimodal LLM with Discrete Sequence Modeling"

Awesome-Unified-Multimodal-Models. Awesome Unified Multimodal Models

metaquery. Official Implementation of Paper Transfer between Modalities with MetaQueries

FluxKits. Python

meanflow. JAX implementation of MeanFlow

UniFork. UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation

OmniFlows. The official implementation of OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows

OmniGen2. OmniGen2: Exploration to Advanced Multimodal Generation. https://arxiv.org/abs/2506.18871

Self-Forcing. Official codebase for "Self Forcing: Bridging Training and Inference in Autoregressive Video Diffusion" (NeurIPS 2025 Spotlight)

Mixture-of-Transformers. Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models. TMLR 2025.

Ming. Ming - facilitating advanced multimodal understanding and generation capabilities built upon the Ling LLM.

DeltaFM. [ICCV 2025] Official Implementation of Contrastive Flow Matching

SimpleAR. Pytorch implementation for the paper titled "SimpleAR: Pushing the Frontier of Autoregressive Visual Generation"

UniWorld. UniWorld: High-Resolution Semantic Encoders for Unified Visual Understanding and Generation

Chain-of-Zoom. [NeurIPS'25 Spotlight] Official repository for "Chain-of-Zoom: Extreme Super-Resolution via Scale Autoregression and Preference Alignment"

eqvae. [ICML'25] EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling.

CFG-Zero-star. Official repo for CFG-Zero*

Diffusion-LLM-Papers. A Collection of Papers on Diffusion Language Models

LLaDA. Official PyTorch implementation for "Large Language Diffusion Models"

Muddit. [ICLR 2026] Official Implementation of Muddit [Meissonic II]: Liberating Generation Beyond Text-to-Image with a Unified Discrete Diffusion Model.

DDO. [ICML 2025 Spotlight] Direct Discriminative Optimization: Reinforcing Diffusion/Autoregressive with GAN Discrimination

Harmon. [ICCV2025]Code Release of Harmonizing Visual Representations for Unified Multimodal Understanding and Generation

Jodi. Jodi: Unification of Visual Generation and Understanding via Joint Modeling

BLIP3o. Official implementation of BLIP3o-Series

LLaDA-V. Python

pytorch-normalizing-flows. Normalizing flows in PyTorch. Current intended use is education not production.

MMaDA. MMaDA - Open-Sourced Multimodal Large Diffusion Language Models (dLLMs with block diffusion, mixed-CoT, unified RL)