Update README.md

README.md

@@ -13,7 +13,12 @@ size_categories:
 - 10K<n<100K
 ---
 
-# UtilityBench: Measuring Language Model Utility When Errors Have Different Costs
+<h1 align="center">
+  <img src="asset/logo.png" alt="" width="36"> UtilityBench
+</h1>
+<p align="center">
+  <strong>Measuring Language Model Utility When Errors Have Different Costs</strong>
+</p>
 
 <p align="left">
   <a href="https://arxiv.org/abs/2610.25924"><img src="https://img.shields.io/badge/arXiv-Paper-b31b1b?logo=arxiv" alt="arXiv"></a>
@@ -47,16 +52,19 @@ The dataset contains 15,000 records in a single `test` split.
 
 UtilityBench builds on 18 source datasets. We thank their authors for making these resources publicly available. Record-level provenance is preserved in `source_id`.
 
-<!-- ## Citations
+## Citations
 
-If you find this work useful, please consider citing:
+If you find this work useful, please consider citing our paper:
 
-```bibtex
-@misc{utilitybench,
-  title={UtilityBench},
-  url={https://github.com/LQF39466/UtilityBench},
+```
+@misc{lei2026utilitybench,
+  author        = {Qifeng Lei and Zhiyong Yang and Zitai Wang and Shilong Bao and Linchao Zhang and Qianqian Xu and Qingming Huang},
+  title         = {{UtilityBench}: Measuring Language Model Utility When Errors Have Different Costs},
+  year          = {2026},
+  eprint        = {2610.25924},
+  archiveprefix = {arXiv},
 }
-``` -->
+```
 
 ## Contact