Update README.md

README.md

@@ -15,4 +15,5 @@ The Rho-1 series are pretrained language models that utilize Selective Language
 In math reasoning pretraining, SLM improves average few-shot accuracy on GSM8k and MATH by over 16%, achieving the baseline performance 5-10x faster.
 
 
-For more details please check our [github](https://github.com/microsoft/rho).
+For more details please check our [github](https://github.com/microsoft/rho) and [paper](https://arxiv.org/abs/2404.07965).
+