Add HyperAlign-Bench dataset card
Browse files
README.md
CHANGED
|
@@ -1,3 +1,76 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
pretty_name: HyperAlign-Bench
|
| 4 |
+
tags:
|
| 5 |
+
- hypergraph
|
| 6 |
+
- large-language-models
|
| 7 |
+
- qwen3
|
| 8 |
+
- hyper-align
|
| 9 |
+
- hypergraph-as-language
|
| 10 |
+
- hyperalign-bench
|
| 11 |
+
- arxiv:2605.21858
|
| 12 |
---
|
| 13 |
+
|
| 14 |
+
# HyperAlign-Bench
|
| 15 |
+
|
| 16 |
+
This repository contains the released **HyperAlign-Bench** data bundle for the paper **Hypergraph as Language**.
|
| 17 |
+
|
| 18 |
+
HyperAlign-Bench is a benchmark for evaluating high-order association modeling in hypergraph-language alignment. It provides vertex classification and hyperedge classification tasks under the same question-answering protocol used by **Hyper-Align**.
|
| 19 |
+
|
| 20 |
+
The released bundle includes processed hypergraph tensors, task samples, prebaked samples, Qwen3-Embedding-0.6B features, and overview features for five datasets.
|
| 21 |
+
|
| 22 |
+
## Files
|
| 23 |
+
|
| 24 |
+
```text
|
| 25 |
+
arxiv_hg/
|
| 26 |
+
cora_cc/
|
| 27 |
+
pubmed/
|
| 28 |
+
dblp/
|
| 29 |
+
imdb/
|
| 30 |
+
```
|
| 31 |
+
|
| 32 |
+
Each dataset directory follows this structure:
|
| 33 |
+
|
| 34 |
+
```text
|
| 35 |
+
processed_data.pt
|
| 36 |
+
meta.json
|
| 37 |
+
samples/
|
| 38 |
+
embeddings/qwen3emb_0.6b/
|
| 39 |
+
overview/qwen3emb_0.6b/
|
| 40 |
+
```
|
| 41 |
+
|
| 42 |
+
- `processed_data.pt` contains the processed hypergraph data.
|
| 43 |
+
- `meta.json` contains dataset metadata.
|
| 44 |
+
- `samples/` contains the VC and HEC task samples.
|
| 45 |
+
- `embeddings/qwen3emb_0.6b/` contains node and hyperedge text features.
|
| 46 |
+
- `overview/qwen3emb_0.6b/` contains overview features used by the released Hyper-Align pipeline.
|
| 47 |
+
|
| 48 |
+
For code, setup, and evaluation instructions, see the GitHub repository:
|
| 49 |
+
|
| 50 |
+
https://github.com/Mengqi-Lei/Hypergraph-as-Language
|
| 51 |
+
|
| 52 |
+
## License
|
| 53 |
+
|
| 54 |
+
The HyperAlign-Bench release is distributed under the Apache License 2.0.
|
| 55 |
+
|
| 56 |
+
Qwen3-Embedding-0.6B is not redistributed here as a model. Users must comply with the license and terms of the upstream model repository:
|
| 57 |
+
|
| 58 |
+
- `Qwen/Qwen3-Embedding-0.6B`
|
| 59 |
+
|
| 60 |
+
## Citation
|
| 61 |
+
|
| 62 |
+
If you use this dataset, please cite:
|
| 63 |
+
|
| 64 |
+
```bibtex
|
| 65 |
+
@article{lei2026hypergraph,
|
| 66 |
+
title={Hypergraph as Language},
|
| 67 |
+
author={Lei, Mengqi and Xie, Guohuan and Ying, Shihui and Du, Shaoyi and Yong, Jun-Hai and Li, Siqi and Gao, Yue},
|
| 68 |
+
journal={arXiv preprint arXiv:2605.21858},
|
| 69 |
+
year={2026}
|
| 70 |
+
}
|
| 71 |
+
```
|
| 72 |
+
|
| 73 |
+
## Links
|
| 74 |
+
|
| 75 |
+
- Paper: https://arxiv.org/abs/2605.21858
|
| 76 |
+
- Code: https://github.com/Mengqi-Lei/Hypergraph-as-Language
|