Refine task categories and tags

This PR refines the `task_categories` metadata to better reflect the primary task of the VCRBench dataset, which is video classification focused on causal reasoning. Some less relevant tags have also been removed for clarity. The description has been slightly improved for better clarity and conciseness.

README.md

@@ -2,10 +2,11 @@
 language:
 - en
 license: mit
-pretty_name: VCRBench
+size_categories:
+- n<1K
 task_categories:
-- video-text-to-text
-- visual-question-answering
+- video-classification
+pretty_name: VCRBench
 tags:
 - video
 - multimodal
@@ -14,15 +15,11 @@ tags:
 - multi-step-reasoning
 - long-form-reasoning
 - large-video-language-model
-- large-multimodal-model
-- multimodal-large-language-model
-size_categories:
-- n<1K
 configs:
 - config_name: default
   data_files:
   - split: test
-    path: "data.json"
+    path: data.json
 ---
 
 # VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models
@@ -36,42 +33,37 @@ configs:
 
 Authors: [Pritam Sarkar](https://pritamsarkar.com) and [Ali Etemad](https://www.aiimlab.com/ali-etemad)
 
-This repository provides the official implementation of **[VCRBench](https://arxiv.org/abs/2505.08455)**.
+This repository provides the official implementation of **[VCRBench](https://arxiv.org/abs/2505.08455)**.  VCRBench is a benchmark dataset for evaluating the causal reasoning capabilities of Large Video Language Models (LVLMs) in visually grounded, goal-driven scenarios. It consists of procedural videos with shuffled steps, requiring LVLMs to identify, reason about, and correctly sequence events to achieve a specific goal.
 
 ![](./sample.png)
 
 ## Usage
 
-Please check our GitHub repo for the details of usage: [VCRBench](https://github.com/pritamqu/VCRBench)
+For detailed usage instructions, please refer to the GitHub repository: [VCRBench](https://github.com/pritamqu/VCRBench)
 
-```
+A basic example:
+
+```python
 from dataset import VCRBench
-dataset=VCRBench(question_file="data.json", 
-                video_root="./",
-                mode='default', 
-                )
-    
+dataset = VCRBench(question_file="data.json", video_root="./", mode='default')
 for sample in dataset:
-    print(sample['question'], )
-    print(sample['answer'], )
+    print(sample['question'])
+    print(sample['answer'])
     print('*'*10)
-
     break
-
 ```
 
 
 ### Licensing Information
 
-This dataset incorporates samples from [CrossTask](https://github.com/DmZhukov/CrossTask/blob/master/LICENSE) that are subject to their respective original licenses. Users must adhere to the terms and conditions specified by these licenses.
-This project does not impose any additional constraints beyond those stipulated in the original licenses. Users must ensure their usage complies with all applicable laws and regulations.
-This repository is released under the **MIT**. See [LICENSE](LICENSE) for details.
+This dataset incorporates samples from [CrossTask](https://github.com/DmZhukov/CrossTask/blob/master/LICENSE) and is subject to their respective original licenses. This repository is released under the **MIT License**. See [LICENSE](LICENSE) for details. Users must adhere to the terms and conditions specified by these licenses.
+
 
 ### Citation Information
 
 If you find this work useful, please use the given bibtex entry to cite our work:
 
-```
+```bibtex
 @misc{sarkar2025vcrbench,
       title={VCRBench: Exploring Long-form Causal Reasoning Capabilities of Large Video Language Models}, 
       author={Pritam Sarkar and Ali Etemad},