Update dataset card

This PR updates the dataset card to include the paper link and change the task category to `question-answering`.

README.md

@@ -1,21 +1,21 @@
 ---
-license: mit
-task_categories:
-- text-generation
 language:
 - en
 - zh
+license: mit
+size_categories:
+- n<1K
+task_categories:
+- question-answering
+pretty_name: S1-Bench
 tags:
 - LRM
 - System1
 - fast-thinking
-pretty_name: S1-Bench
-size_categories:
-- n<1K
 ---
 
-The benchmark constructed in paper ***S1-Bench: A Simple Benchmark for Evaluating System 1 Thinking Capability of Large Reasoning Models***.
+The benchmark constructed in paper [S1-Bench: A Simple Benchmark for Evaluating System 1 Thinking Capability of Large Reasoning Models](https://huggingface.co/papers/2504.10368).
 
 S1-Bench is a novel benchmark designed to evaluate Large Reasoning Models' performance on simple tasks that favor intuitive *system 1* thinking rather than deliberative *system 2* reasoning. 
 
-S1-Bench comprises 422 question-answer pairs across four major categories and 28 subcategories, balanced with 220 English and 202 Chinese questions. 
+S1-Bench comprises 422 question-answer pairs across four major categories and 28 subcategories, balanced with 220 English and 202 Chinese questions.