Update README.md

README.md

@@ -9,7 +9,7 @@ In particular, we store the pre-trained vision encoder and ICRT model separately
 
 Please refer to the [code](https://github.com/Max-Fu/icrt) on installing the repo, training and inferencing the model. 
 
-**Dataset Structure**
+## Dataset Structure
 
 ```
 ICRT-MT
@@ -42,4 +42,17 @@ ICRT-MT
 │       ...
 └── merged_data_part1_keys.json
 ...
+```
+
+## Citation
+Please give us a star 🌟 on Github to support us!
+
+Please cite our work if you find our work inspiring or use our code in your work:
+```
+@article{fu2024icrt,
+    title={In-Context Imitation Learning via Next-Token Prediction}, 
+    author={Letian Fu and Huang Huang and Gaurav Datta and Lawrence Yunliang Chen and William Chung-Ho Panitch and Fangchen Liu and Hui Li and Ken Goldberg},
+    journal={arXiv preprint arXiv:2408.15980},
+    year={2024}
+}
 ```