Add paper link and object-detection task category

This PR adds a direct link to the paper on the Hugging Face paper page and includes the `object-detection` task category in the metadata, aligning with the recommended tags for GUI grounding benchmarks.

README.md

@@ -1,8 +1,4 @@
 ---
-license: cc-by-4.0
-task_categories:
-- image-to-text
-- visual-question-answering
 language:
 - en
 - de
@@ -10,6 +6,14 @@ language:
 - es
 - it
 - nl
+license: cc-by-4.0
+size_categories:
+- 1K<n<10K
+task_categories:
+- image-to-text
+- visual-question-answering
+- object-detection
+pretty_name: Pointerbench
 tags:
 - gui-grounding
 - computer-use
@@ -19,13 +23,12 @@ tags:
 - spreadsheets
 - text-grounding
 - professional-apps
-pretty_name: Pointerbench
-size_categories:
-- 1K<n<10K
 ---
 
 # Pointerbench
 
+Paper: [PointerBench: A Diagnostic Benchmark for Fine-Grained GUI Grounding](https://huggingface.co/papers/2610.31274)
+
 Pointerbench is a small GUI grounding benchmark suite for computer-use models.
 Each example has one screenshot, one instruction, target geometry in absolute
 pixels, and a binary evaluation rule.
@@ -174,4 +177,4 @@ See each subset README for the exact distribution, schema details, and examples.
 ## License
 
 Dataset images and annotations are released under CC BY 4.0. The included
-evaluation scripts are released under MIT.
+evaluation scripts are released under MIT.