Add pipeline tag and paper link

This PR adds the missing `pipeline_tag` (`robotics`) to the YAML metadata to improve discoverability of this robotics model. It also adds a direct link to the paper in the model card content.

README.md

@@ -1,15 +1,18 @@
 ---
-license: mit
-library_name: pytorch
 base_model: Wan-AI/Wan2.2-TI2V-5B-Diffusers
+library_name: pytorch
+license: mit
+pipeline_tag: robotics
 tags:
-  - robotics
-  - forward-dynamics
-  - video
-  - scene-flow
-  - lora
+- robotics
+- forward-dynamics
+- video
+- scene-flow
+- lora
 ---
 
+This repository contains the model described in [FloMo: 3D Scene Flow as a World Action Model Intermediate for Learning from Human Video](https://huggingface.co/papers/2611.05062).
+
 # FloMo midtrain base checkpoints
 
 FloMo does any-to-any forward-dynamics cotraining for robot learning from video.
@@ -72,4 +75,4 @@ c50dd53732951f338ed28d34f57aa5f2dabd5da98fe495fc3194b5270655ceaa  va/flomo_va.pt
 ## License
 
 The FloMo code and these weights use the MIT license. The base model
-[Wan2.2-TI2V-5B](https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B) uses the Apache-2.0 license.
+[Wan2.2-TI2V-5B](https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B) uses the Apache-2.0 license.