Improve model card and add robotics pipeline tag

README.md

@@ -1,6 +1,32 @@
 ---
 license: mit
+pipeline_tag: robotics
 ---
-This is the model of the paper [MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction](https://jmsnu.github.io/).
 
-Please refer to https://github.com/jmSNU/mvp_lam for further details of this model
+# MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction
+
+[**Paper**](https://arxiv.org/abs/2602.03668) | [**Project Page**](https://jmsnu.github.io/) | [**Code**](https://github.com/jmSNU/mvp_lam)
+
+MVP-LAM (Multi-ViewPoint Latent Action Model) is a framework designed to learn latent actions that are highly informative about ground-truth actions from multi-view videos. It uses a cross-viewpoint reconstruction objective, ensuring that a latent action from one view can explain the future in another view, which reduces reliance on viewpoint-specific cues and results in more action-centric representations.
+
+## Highlights
+
+- **Cross-viewpoint reconstruction**: Latent actions inferred from one view must explain the future in another view, promoting viewpoint-invariant features.
+- **Action-centric representations**: Achieves higher mutual information with ground-truth actions and improved action prediction, even in out-of-distribution scenarios.
+- **Improved VLA pretraining**: Using MVP-LAM latent actions as pseudo-labels enhances the downstream manipulation performance of Vision-Language-Action (VLA) models on benchmarks like SIMPLER and LIBERO-Long.
+
+## Citation
+
+If you find this work useful, please cite:
+
+```bibtex
+@misc{lee2026mvplam,
+  title     = {MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstruction},
+  author    = {Jung Min Lee and Dohyeok Lee and Seokhun Ju and Taehyun Cho and Jin Woo Koo and Li Zhao and Sangwoo Hong and Jungwoo Lee},
+  year      = {2026},
+  eprint    = {2602.03668},
+  archivePrefix = {arXiv},
+  primaryClass  = {cs.RO},
+  url       = {https://arxiv.org/abs/2602.03668}
+}
+```