Add pipeline tag, library name, and citation to model card

README.md

@@ -1,9 +1,11 @@
 ---
-license: apache-2.0
-datasets:
-- internlm/VC-RewardBench
 base_model:
 - Qwen/Qwen3-VL-8B-Instruct
+datasets:
+- internlm/VC-RewardBench
+license: apache-2.0
+library_name: transformers
+pipeline_tag: image-text-to-text
 ---
 
 <p align="center">
@@ -12,7 +14,7 @@ base_model:
 
 # Visual-ERM
 
-Visual-ERM is a **multimodal generative reward model** for **vision-to-code** tasks.  
+Visual-ERM is a **multimodal generative reward model** for **vision-to-code** tasks, introduced in the paper [Visual-ERM: Reward Modeling for Visual Equivalence](https://huggingface.co/papers/2603.13224).  
 It evaluates outputs directly in the **rendered visual space** and produces **fine-grained**, **interpretable**, and **task-agnostic** discrepancy feedback for structured visual reconstruction.
 
 <p align="center">
@@ -147,7 +149,12 @@ Visual-ERM is intended for:
 If you find this model useful, please consider citing:
 
 ```bibtex
-TBD
+@article{liu2026visualerm,
+  title={Visual-ERM: Reward Modeling for Visual Equivalence},
+  author={Liu, Ziyu and Ding, Shengyuan and Fang, Xinyu and Dai, Xuanlang and Yang, Penghui and Liang, Jianze and Wang, Jiaqi and Chen, Kai and Lin, Dahua and Zang, Yuhang},
+  journal={arXiv preprint arXiv:2603.13224},
+  year={2026}
+}
 ```
 
 ## Contact