Update README.md

README.md

@@ -143,5 +143,14 @@ From these, we compute the **Safety-score** as a composite metric. To assess con
 
 ## Citation
 ```
+@misc{pan2025omnisafetybenchbenchmarksafetyevaluation,
+      title={Omni-SafetyBench: A Benchmark for Safety Evaluation of Audio-Visual Large Language Models}, 
+      author={Leyi Pan and Zheyu Fu and Yunpeng Zhai and Shuchang Tao and Sheng Guan and Shiyu Huang and Lingzhe Zhang and Zhaoyang Liu and Bolin Ding and Felix Henry and Lijie Wen and Aiwei Liu},
+      year={2025},
+      eprint={2508.07173},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2508.07173}, 
+}
 ```