Update BehaviorBench paper citation

README.md

@@ -30,6 +30,9 @@ solving (IEO). All examples are released as chat-formatted
 This repository hosts the **evaluation data only**. The benchmark code (loaders,
 prompts, metrics) is released separately from this dataset.
 
+Public dataset URL: https://huggingface.co/datasets/befm/BehaviorBench
+Paper: https://arxiv.org/abs/2606.24162
+
 ## Subsets
 
 | Subset | Task name | Files | Rows |
@@ -138,9 +141,13 @@ behavior trigger a new minor version (current: `1.0.0`).
 ## Citation
 
 ```bibtex
-@misc{behaviorbench2026,
-  title  = {BehaviorBench: Benchmarking Foundation Models for Behavioral Science Tasks},
-  author = {BeFM Authors},
-  year   = {2026}
+@misc{huang2026behaviorbenchbenchmarkingfoundationmodels,
+  title={BehaviorBench: Benchmarking Foundation Models for Behavioral Science Tasks},
+  author={Jin Huang and Yutong Xie and Wanli Song and Xingjian Zhang and Walter Yuan and Matthew O. Jackson and Qiaozhu Mei},
+  year={2026},
+  eprint={2606.24162},
+  archivePrefix={arXiv},
+  primaryClass={cs.CL},
+  url={https://arxiv.org/abs/2606.24162}
 }
 ```