Add link to paper

This PR adds a link to the paper introducing UnCoBench, "Beyond Perfect Teammates: Multi-Agent Collaboration with Unreliable Colleagues" (https://huggingface.co/papers/2610.16117), in the dataset card for easier discoverability. No metadata changes are needed; the existing metadata already correctly includes task categories, license, size, and tags.

README.md

@@ -1,12 +1,12 @@
 ---
-license: cc-by-4.0
 language:
 - en
-pretty_name: UnCoBench
+license: cc-by-4.0
 size_categories:
 - 1K<n<10K
 task_categories:
 - text-generation
+pretty_name: UnCoBench
 tags:
 - text
 - benchmark
@@ -19,7 +19,7 @@ tags:
 
 # UnCoBench
 
-[GitHub Repository](https://github.com/iNLP-Lab/UnCoBench)
+[GitHub Repository](https://github.com/iNLP-Lab/UnCoBench) | [Paper](https://huggingface.co/papers/2610.16117)
 
 UnCoBench evaluates how well a main agent completes collaborative tasks when other agents exhibit unreliable behavior. The dataset contains **1,000 groups of paired English tasks** across **five settings**, yielding **5,000 task instances**.
 
@@ -48,4 +48,4 @@ The four primary settings support paired comparisons of individual unreliable be
 
 ## License
 
-The benchmark data is licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/), consistent with the [data license](https://github.com/iNLP-Lab/UnCoBench/blob/main/LICENSE-DATA). When redistributing the data, credit the UnCoBench contributors, include the license, and indicate any modifications.
+The benchmark data is licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/), consistent with the [data license](https://github.com/iNLP-Lab/UnCoBench/blob/main/LICENSE-DATA). When redistributing the data, credit the UnCoBench contributors, include the license, and indicate any modifications.