Add task categories and link to paper/code

README.md

@@ -1,4 +1,11 @@
 ---
+language:
+- en
+license: apache-2.0
+size_categories:
+- n<1K
+task_categories:
+- text-generation
 dataset_info:
   features:
   - name: source
@@ -63,16 +70,13 @@ configs:
     path: data/dev-*
   - split: test
     path: data/test-*
-license: apache-2.0
-language:
-- en
-size_categories:
-- n<1K
 ---
 
+# Math-IF Dataset Card
 
+This dataset is associated with the paper [From Leaky Thoughts to Private Reasoning: Controlling What LRMs Say to Themselves](https://huggingface.co/papers/2602.24210).
 
-# Math-IF Dataset Card
+The official code repository for the project is available here: [UKPLab/arxiv2026-controllable-reasoning-models](https://github.com/UKPLab/arxiv2026-controllable-reasoning-models).
 
 ## Dataset Description
 
@@ -87,12 +91,11 @@ In this repository, Math-IF is used as both a development set and a test benchma
 
 - Evaluate how well models follow explicit instructions when solving math problems.
 
-
 The dataset is intended for **research and benchmarking** only.
 
 ## Dataset Structure
 
-From the accompanying paper in this repository (see `paper/`):
+From the accompanying paper:
 
 - **Size**:
   - **Dev**: 90 examples  
@@ -131,7 +134,6 @@ Math-IF was introduced to study the trade-off between reasoning performance and
 
 ## Citation
 
-
 ```bibtex
 @misc{puerto2026controllablereasoningmodelsprivate,
       title={Controllable Reasoning Models Are Private Thinkers},