Update README.md

README.md

@@ -32,7 +32,7 @@ configs:
   <h3 align="center">Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents</h3>
 
   <p align="center">
-    <a href="#">📄 arXiv</a> •
+    <a href="https://arxiv.org/abs/2608.12764">📄 arXiv</a> •
     <a href="https://github.com/hkust-nlp/SSPO">💻 Code</a> •
     <a href="https://huggingface.co/datasets/WaitHZ/SSPO-data">🤗 Dataset</a>
   </p>
@@ -74,7 +74,15 @@ hf download WaitHZ/SSPO-data \
 ## 🚩Citation
 
 ```bib
-
+@misc{wu2026outcomerewardssteplevelselfdistilled,
+      title={Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents}, 
+      author={Haoze Wu and Chuqiao Kuang and Tianyi Zhuang and Xiaoguang Li},
+      year={2026},
+      eprint={2608.12764},
+      archivePrefix={arXiv},
+      primaryClass={cs.LG},
+      url={https://arxiv.org/abs/2608.12764}, 
+}
 ```
 
 ## 🌻Acknowledgement