Add pipeline tag and links to paper, project, and code

This PR adds the missing `pipeline_tag: any-to-any` to the model card metadata so the model is correctly classified as a unified multimodal model on the Hub. It also adds links to the paper, project page, and GitHub repository in the README content for easier access and better documentation.

README.md

@@ -1,20 +1,25 @@
 ---
+base_model: Qwen/Qwen3-8B
 license: other
 license_name: nvidia-one-way-noncommercial-license
 license_link: https://developer.download.nvidia.com/licenses/NVIDIA-OneWay-Noncommercial-License-22Mar2022.pdf
-base_model: Qwen/Qwen3-8B
+pipeline_tag: any-to-any
 tags:
-  - multimodal
-  - image-generation
-  - video-generation
-  - image-to-text
-  - video-text-to-text
+- multimodal
+- image-generation
+- video-generation
+- image-to-text
+- video-text-to-text
 ---
 
 # PixelUMM
 
 **PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation**
 
+**Paper:** [PixelUMM: Encoder-Free Unified Image and Video Understanding and Generation](https://huggingface.co/papers/2609.38597)  
+**Project:** [PixelUMM Project Page](https://nv-tlabs.github.io/PixelUMM/)  
+**Code:** [GitHub](https://github.com/nv-tlabs/PixelUMM)
+
 PixelUMM is an NVIDIA-developed, encoder-free unified multimodal model for joint
 understanding and generation across text, images, and video directly in pixel
 space.
@@ -101,4 +106,4 @@ complying with all applicable upstream licenses and terms.
 ## References
 
 - [Qwen3-8B model repository](https://huggingface.co/Qwen/Qwen3-8B)
-- [BAGEL source repository](https://github.com/ByteDance-Seed/Bagel)
+- [BAGEL source repository](https://github.com/ByteDance-Seed/Bagel)