Update README.md

README.md

@@ -81,7 +81,7 @@ Using RefOI, we evaluate several state-of-the-art VLMs and uncover **three tiers
 ![Overview](vlm-reg.png)
 
 
-## ## Dataset Schema and Split
+## Dataset Schema and Split
 
 ### Data Fields
 Each entry in the dataset contains the following fields:
@@ -163,7 +163,7 @@ If you find our data useful, please consider citing our work:
 @misc{vlmreg2025,
   title        = {Vision–Language Models Are Not Pragmatically Competent in Referring Expression Generation},
   year         = {2025},
-  note         = {Under review at COLM 2025, ACL GEM 2025, CVinW CVPR 2025},
+  note         = {Under review at COLM 2025, ACL GEM 2025},
   url          = {https://vlm-reg.github.io},
 }
 ```