Update dataset card with task category, paper link, and usage instructions

Hi! I'm Niels from the community science team at Hugging Face. I've updated the dataset card for OBJ3D to improve its discoverability and documentation. Changes include:
- Adding the `image-to-video` task category.
- Linking the dataset to the paper [Latent Particle World Models](https://huggingface.co/papers/2603.04553).
- Adding links to the official GitHub repository and project page.
- Including sample usage commands for training models on this dataset as documented in the GitHub README.
- Adding the BibTeX citation.

README.md

@@ -1,5 +1,7 @@
 ---
 license: cc-by-4.0
+task_categories:
+- image-to-video
 tags:
 - simulation
 - clevr
@@ -8,5 +10,39 @@ tags:
 - video
 - prediction
 ---
-OBJ3D dataset from G-SWM (https://github.com/zhixuan-lin/G-SWM).
-Contains video frames of synthetic CLEVR-like objects colliding.
+
+# OBJ3D Dataset
+
+OBJ3D dataset originally from [G-SWM](https://github.com/zhixuan-lin/G-SWM). It contains video frames of synthetic CLEVR-like objects colliding.
+
+This dataset is used for evaluating object-centric world models and stochastic dynamics, as featured in the paper: **[Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling](https://huggingface.co/papers/2603.04553)**.
+
+- **Project Page:** [https://taldatech.github.io/lpwm-web](https://taldatech.github.io/lpwm-web)
+- **GitHub Repository:** [https://github.com/taldatech/lpwm](https://github.com/taldatech/lpwm)
+
+## Sample Usage
+
+To train models on this dataset using the official implementation, you can use the following commands:
+
+### Single-GPU Training (DLPv3)
+```bash
+python train_dlp.py --dataset obj3d_img
+```
+
+### Multi-GPU Training (LPWM)
+```bash
+accelerate launch --config_file ./accel_conf.yml train_lpwm_accelerate.py --dataset obj3d128
+```
+
+## Citation
+
+```bibtex
+@inproceedings{
+daniel2026latent,
+title={Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling},
+author={Tal Daniel and Carl Qi and Dan Haramati and Amir Zadeh and Chuan Li and Aviv Tamar and Deepak Pathak and David Held},
+booktitle={The Fourteenth International Conference on Learning Representations},
+year={2026},
+url={https://openreview.net/forum?id=lTaPtGiUUc}
+}
+```