Update pipeline tag and add paper/code links

Hi! I'm Niels from the community science team at Hugging Face.

This PR improves your model card by:
- Updating the `pipeline_tag` to `graph-ml` to better reflect the model's domain in protein structure and geometric deep learning.
- Adding a link to the paper: [Protein Fold Classification at Scale: Benchmarking and Pretraining](https://huggingface.co/papers/2605.18552).
- Adding a link to the official [GitHub repository](https://github.com/BorgwardtLab/TEDBench).

These changes help users discover the model more easily and provide direct access to the associated research and code.

README.md

@@ -1,26 +1,26 @@
 ---
 library_name: tedbench
-tags:
-  - protein
-  - structure
-  - fold-classification
-  - tedbench
-pipeline_tag: feature-extraction
 license: bsd-3-clause
+pipeline_tag: graph-ml
+tags:
+- protein
+- structure
+- fold-classification
+- tedbench
 ---
 
 # TEDBench — Pretrained autoencoder (structure + sequence)
 
 **Variant:** `miae_b` + seq  |  **Parameters:** 102M  |  **Layers:** 12  |  **Hidden dim:** 768  |  **Attn heads:** 12
 
-This is a **pretrained MiAE** checkpoint. Use it as a feature extractor or as the starting point for fine-tuning.
+This repository contains the **pretrained MiAE** checkpoint introduced in the paper [Protein Fold Classification at Scale: Benchmarking and Pretraining](https://huggingface.co/papers/2605.18552). Use it as a feature extractor or as the starting point for fine-tuning.
+
+- **Paper:** [Protein Fold Classification at Scale: Benchmarking and Pretraining](https://huggingface.co/papers/2605.18552)
+- **Repository:** [BorgwardtLab/TEDBench](https://github.com/BorgwardtLab/TEDBench)
 
 > **+seq variant** — sequence embeddings are concatenated to the geometric encoder input (`model.use_seq_input=true`).
 
-Part of the [TEDBench](https://github.com/BorgwardtLab/TEDBench) benchmark for
-protein fold classification (ICML 2026). MiAE is an SE(3)-invariant masked
-autoencoder that masks up to 90% of backbone frames and reconstructs the full
-structure with a lightweight decoder.
+Part of the [TEDBench](https://github.com/BorgwardtLab/TEDBench) benchmark for protein fold classification (ICML 2026). MiAE is an SE(3)-invariant masked autoencoder that masks up to 90% of backbone frames and reconstructs the full structure with a lightweight decoder.
 
 ## Architecture sizes
 
@@ -61,4 +61,4 @@ model.eval()
   booktitle={Proceedings of the 43rd International Conference on Machine Learning},
   year={2026}
 }
-```
+```