Semantics Lead the Way: Harmonizing Semantic and Texture Modeling with Asynchronous Latent Diffusion

PaCoRe: Learning to Scale Test-Time Compute with Parallel Coordinated Reasoning

Tiny AutoEncoder for Hunyuan Video (and other video models)

Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation

SAM 3D Objects

🚀🚀 Efficient implementations of Native Sparse Attention

LongLive: Real-time Interactive Long Video Generation

rCM: SOTA Diffusion Distillation & Few-Step Video Generation based on sCM/MeanFlow

Cosmos-Predict2.5, the latest version of the Cosmos World Foundation Models (WFMs) family, specialized for simulating and predicting the future state of the world in the form of video.

🚁🚀 2025机场推荐 | 熊猫VPN（PandaVPNPro）已确定跑路！快连VPN ，小牛加速器（小牛VPN）体验不佳且价格贵。低价机场，便宜机场，平价机场，性价比机场，高速机场，稳定机场，一元机场，翻墙机场，付费机场，收费机场，冷门机场，优质机场，廉价机场。机场评测，中国翻墙，科学上网，梯子。非永久免费梯子，官网，非永久免费VPN，非免费机场！VPN China。按量计费机场，按量付费…

A unified inference and post-training framework for accelerated video generation.

Sampling profiler for Python programs

Native Multimodal Models are World Learners

"ViMax: Agentic Video Generation (Director, Screenwriter, Producer, and Video Generator All-in-One)"

Unofficial extension implementation of Self-Forcing to support I2V && 14B training.

Large Language Model (LLM) Systems Paper List

Official training code for MUG-V 10B video generation model. Built on Megatron-LM (v0.14.0) with production-ready distributed training for 10B DiT.

Contexts Optical Compression

Krea Realtime 14B. An open-source realtime AI video model.

pi-Flow: Policy-Based Few-Step Generation via Imitation Distillation

Accelerate inference without tears

A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.

Infinite-Forcing: Towards Infinite-Long Video Generation

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

The best ChatGPT that $100 can buy.

A comprehensive list of papers for the definition of World Models and using World Models for General Video Generation, Embodied AI, and Autonomous Driving, including papers, codes, and related webs…