Add model card, pipeline tag, and links to paper/code

README.md

@@ -1,7 +1,76 @@
 ---
-license: mit
 base_model:
 - alibaba-pai/Wan2.1-Fun-14B-InP
+license: mit
+pipeline_tag: image-to-video
 tags:
 - World
----
+---
+
+# EgoSim: Egocentric World Simulator for Embodied Interaction Generation
+
+This repository contains the weights for **EgoSim-14B**, a closed-loop egocentric world simulator presented in the paper [EgoSim: Egocentric World Simulator for Embodied Interaction Generation](https://huggingface.co/papers/2604.01001).
+
+[**Project Page**](https://egosimulator.github.io/) | [**GitHub Code**](https://github.com/jinkun-hao/EgoSim) | [**Paper**](https://huggingface.co/papers/2604.01001)
+
+![Teaser](https://raw.githubusercontent.com/jinkun-hao/EgoSim/main/images/teaser.png)
+
+## Overview
+**EgoSim** is an egocentric world simulator for embodiment interaction generation. Given an initial 3D state and a sequence of actions, EgoSim generates temporally and spatially consistent egocentric observations with high-quality dexterous interactions. EgoSim also persistently updates a 3D scene state for continuous simulation.
+
+## Quickstart
+
+### Installation
+
+Requires Python 3.10+ and CUDA 12.1+.
+
+```bash
+git clone https://github.com/jinkun-hao/EgoSim.git
+cd EgoSim
+
+conda create -n egosim python=3.10 -y
+conda activate egosim
+
+# Install PyTorch
+pip install torch torchvision
+# Install flash attention
+pip install flash-attn --no-build-isolation
+pip install -r requirements.txt
+```
+
+### Model Download
+
+Download the **EgoSim-14B** weights into the project root:
+
+```bash
+huggingface-cli download wuzhi-hao/EgoSim --local-dir ./EgoSim-14B
+```
+
+### Inference
+
+Download the demo samples from the [official repository instructions](https://github.com/jinkun-hao/EgoSim), extract under `tests/samples/`, and run:
+
+```bash
+# Egodex — quick test with bundled mini samples
+PYTHONPATH=. python egowm/inference/runner.py \
+  --dataset egodex \
+  --model_root ./EgoSim-14B \
+  --dataset_root tests/samples/demo_data/egodex \
+  --metadata_path tests/samples/demo_data/egodex_metadata.csv \
+  --output_dir output_egodex \
+  --num_inference_steps 50 \
+  --gpu_id 0
+```
+
+## Citation
+
+If you find EgoSim useful in your research, please cite the paper:
+
+```bibtex
+@article{hao2026egosim,
+  title={EgoSim: Egocentric World Simulator for Embodied Interaction Generation},
+  author={Hao, Jinkun and Jia, Mingda and Wang, Ruiyan and Liu, Xihui and Yi, Ran and Ma, Lizhuang and Pang, Jiangmiao and Xu, Xudong},
+  journal={arXiv preprint arXiv:2604.01001},
+  year={2026}
+}
+```