Add model card metadata and paper link

README.md

@@ -1,13 +1,27 @@
 ---
-license: apache-2.0
 base_model:
 - meta-llama/Meta-Llama-3-8B-Instruct
+license: apache-2.0
+library_name: transformers
+pipeline_tag: text-generation
 ---
+
+# Mix Data or Merge Models? (RESM)
+
+This repository contains the model weights for the 3H-aligned LLM (balancing Helpfulness, Honesty, and Harmlessness) developed in the paper [Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging](https://huggingface.co/papers/2502.06876).
+
+## Description
+This model is constructed using a novel merging method called **RESM** (**R**eweighting **E**nhanced task **S**ingular **M**erging). RESM addresses challenges like preference noise accumulation and layer sparsity adaptation through outlier weighting and sparsity-aware rank selection strategies. The approach aims to achieve a balanced alignment across the 3H dimensions (Helpfulness, Honesty, and Harmlessness) more effectively than traditional data mixture or standard model merging techniques.
+
+- **Base Model:** [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)
+- **Paper:** [Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging](https://huggingface.co/papers/2502.06876)
+
 ## Citation
-```
+```bibtex
 @article{yang2025mix,
   title={Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging},
   author={Yang, Jinluan and Jin, Dingnan and Tang, Anke and Shen, Li and Zhu, Didi and Chen, Zhengyu and Wang, Daixin and Cui, Qing and Zhang, Zhiqiang and Zhou, Jun and others},
   journal={arXiv preprint arXiv:2502.06876},
   year={2025}
-}
+}
+```