DriveDreamer-Policy: A Geometry-Grounded World–Action Model for Unified Generation and Planning

LDA-1B: Scaling Latent Dynamics Action Model via Universal Embodied Data Ingestion

[Ultra Powerful Few-Step Diffusion RL] TDM-R1: Reinforcing Few-Step Diffusion Models with Non-Differentiable Reward

Helios: Real Real-Time Long Video Generation Model

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

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

Visualization toolkit for Pi3DET-Dataset

TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

[CVPR 2026 Oral] WorldLens: Full-Spectrum Evaluations of Driving World Models in Real World

[ICLR 2026] Official Implementation of "UniSplat: Unified Spatio-Temporal Fusion via 3D Latent Scaffolds for Dynamic Driving Scene Reconstruction""

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

Let's finetune video generation models!

[NeurIPS 2025 DB Track] 3EED: Ground Everything Everywhere in 3D

[NeurIPS 2025] Deep Memory Backtracking for Long Video Understanding

[NeurIPS2025] ReID5o: Achieving Omni Multi-modal Person Re-identification in a Single Model

A high-throughput and memory-efficient inference and serving engine for LLMs

[ICLR 2026] LongLive: Real-time Interactive Long Video Generation

[ICLR 2026] Lyra: Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation

Making large AI models cheaper, faster and more accessible

Discover Unknown Unsafe Events via Generative Simulation

LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

🌐 Vision-Language-Action Models for Autonomous Driving: Past, Present, and Future

Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.5, DeepSeek, gpt-oss locally.

Enjoy the magic of Diffusion models!

[CoRL '25] Pseudo-Simulation for Autonomous Driving; [NeurIPS '24] NAVSIM: Data-Driven Non-Reactive Autonomous Vehicle Simulation and Benchmarking

[CoRLW 2025 (Oral), IASEAI 2026] Implementation for "Challenger: Affordable Adversarial Driving Video Generation"

Cosmos-Predict2 is a collection of general-purpose world foundation models for Physical AI that can be fine-tuned into customized world models for downstream applications.

[ICCV 2025] Official code of "ORION: A Holistic End-to-End Autonomous Driving Framework by Vision-Language Instructed Action Generation"

Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)