Update README.md

README.md

@@ -18,7 +18,7 @@ tags:
 ![Benchmark scatter: FLOPs and tokens vs benchmark average for HRM-Text-1B vs comparable models](benchmark_scatter.png)
 
 <p align="center">
-  <a href="https://sapientinc.github.io/HRM-Text/assets/HRM_Text.pdf"><img src="https://img.shields.io/badge/Paper-PDF-red" alt="Paper"></a>
+  <a href="https://arxiv.org/pdf/2605.20613"><img src="https://img.shields.io/badge/Paper-arXiv-red?logo=arxiv&logoColor=white" alt="arXiv Paper"></a>
   <a href="https://github.com/sapientinc/HRM-Text"><img alt="GitHub" src="https://img.shields.io/badge/GitHub-sapientinc%2FHRM--Text-181717?logo=github&logoColor=white"></a>
 </p>
 
@@ -148,4 +148,16 @@ Pre-trained on a sampled mixture of publicly available text corpora. The full da
 
 ## Citation
 
-Citation information will be added with the accompanying paper.
+If you find this project or our paper useful, please consider citing our paper:
+
+```
+@misc{wang2026hrmtextefficientpretrainingscaling,
+      title={HRM-Text: Efficient Pretraining Beyond Scaling}, 
+      author={Guan Wang and Changling Liu and Chenyu Wang and Cai Zhou and Yuhao Sun and Yifei Wu and Shuai Zhen and Luca Scimeca and Yasin Abbasi Yadkori},
+      year={2026},
+      eprint={2605.20613},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL},
+      url={https://arxiv.org/abs/2605.20613}, 
+}
+```