Update README.md

README.md

@@ -23,7 +23,7 @@ LLaSE-G1 is a unified speech enhancement model capable of handling multiple task
 
 To mitigate acoustic inconsistency, LLaSE-G1 employs continuous representations from **WavLM** as input and predicts speech tokens using **X-Codec2**, maximizing acoustic preservation. The model surpasses prior task-specific discriminative and generative speech enhancement models, demonstrating scaling effects at test time and emerging capabilities for unseen speech enhancement tasks.
 
-For more details, refer to our paper: [LLaSE-G1 Paper](https://submission-papers.github.io/LLaSE-G1-demo-page/)
+For more details, refer to our paper: [LLaSE-G1 Paper](https://arxiv.org/abs/2503.00493)
 
 ## Demo
 
@@ -93,6 +93,15 @@ Our pretrained model is available on [Hugging Face](https://huggingface.co/ASLP-
 
 If you find this work useful, please cite our paper:
 
+> @misc{kang2025llaseg1incentivizinggeneralizationcapability,
+      title={LLaSE-G1: Incentivizing Generalization Capability for LLaMA-based Speech Enhancement}, 
+      author={Boyi Kang and Xinfa Zhu and Zihan Zhang and Zhen Ye and Mingshuai Liu and Ziqian Wang and Yike Zhu and Guobin Ma and Jun Chen and Longshuai Xiao and Chao Weng and Wei Xue and Lei Xie},
+      year={2025},
+      eprint={2503.00493},
+      archivePrefix={arXiv},
+      primaryClass={eess.AS},
+      url={https://arxiv.org/abs/2503.00493}, 
+}
 
 ## License