A Curated List of Awesome Works in World Modeling, Aiming to Serve as a One-stop Resource for Researchers, Practitioners, and Enthusiasts Interested in World Modeling.

Native Multimodal Models are World Learners

Detect Anything via Next Point Prediction (Based on Qwen2.5-VL-3B)

Official PyTorch Implementation of "Diffusion Transformers with Representation Autoencoders"

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

Recipes to train reward model for RLHF.

[NeurIPS 2025] Pixel-Perfect Depth

Thinking with Videos from Open-Source Priors. We reproduce chain-of-frames visual reasoning by fine-tuning open-source video models. Give it a star 🌟 if you find it useful.

UniVid: The Open-Source Unified Video Model

Fully Open Framework for Democratized Multimodal Training

Code of BRIDGE: Building Reinforcement-Learning Depth-to-Image Data Generation Engine for Monocular Depth Estimation

A minimal implementation of DeepMind's Genie world model

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

Qwen3-VL is the multimodal large language model series developed by Qwen team, Alibaba Cloud.

Qwen3-omni is a natively end-to-end, omni-modal LLM developed by the Qwen team at Alibaba Cloud, capable of understanding text, audio, images, and video, as well as generating speech in real time.

OmniWorld: A Multi-Domain and Multi-Modal Dataset for 4D World Modeling

Official repository for the UAE paper, unified-GRPO, and unified-Bench

[NeurIPS 2025 (Spotlight)] The implementation for the paper "4DGT Learning a 4D Gaussian Transformer Using Real-World Monocular Videos"

Fine-tuning & Reinforcement Learning for LLMs. 🦥 Train OpenAI gpt-oss, DeepSeek-R1, Qwen3, Gemma 3, TTS 2x faster with 70% less VRAM.

slime is an LLM post-training framework for RL Scaling.

Unified Reinforcement Learning Framework

An Easy-to-use, Scalable and High-performance RLHF Framework based on Ray (PPO & GRPO & REINFORCE++ & vLLM & Ray & Dynamic Sampling & Async Agentic RL)

Extend OpenRLHF to support LMM RL training for reproduction of DeepSeek-R1 on multimodal tasks.

A fork to add multimodal model training to open-r1

Witness the aha moment of VLM with less than $3.

Fully open reproduction of DeepSeek-R1

Train transformer language models with reinforcement learning.

EasyR1: An Efficient, Scalable, Multi-Modality RL Training Framework based on veRL