Add DyBench dataset card and link to paper

Hi, I'm Niels from the community science team at Hugging Face.

This PR improves the dataset card for DyBench by:
- Adding the `video-text-to-text` task category to the metadata.
- Linking the research paper: [Learning Spatiotemporal Sensitivity in Video LLMs via Counterfactual Reinforcement Learning](https://huggingface.co/papers/2605.21988).
- Adding links to the project page and GitHub repository.
- Providing a description of the benchmark and its categories (Reversible Dynamics, Moving Direction, and Event Sequence).

README.md

@@ -1,5 +1,9 @@
 ---
+language:
+- en
 license: mit
+task_categories:
+- video-text-to-text
 extra_gated_fields:
   Name: text
   Company/Organization: text
@@ -15,6 +19,26 @@ configs:
   data_files: json/moving_direction.json
 - config_name: reversible_dynamics
   data_files: json/reversible_dynamics.json
-language:
-- en
----
+---
+
+# DyBench
+
+[**Project Page**](https://ddz16.github.io/crpo.github.io/) | [**Paper**](https://huggingface.co/papers/2605.21988) | [**GitHub**](https://github.com/ddz16/CRPO)
+
+DyBench is a paired counterfactual video benchmark introduced in the paper "[Learning Spatiotemporal Sensitivity in Video LLMs via Counterfactual Reinforcement Learning](https://huggingface.co/papers/2605.21988)". 
+
+The benchmark is designed to evaluate the **spatiotemporal sensitivity** of Video Large Language Models (Video LLMs). It addresses the issue of models relying on "shortcuts" (such as single-frame cues or language priors) rather than tracking actual video dynamics. DyBench utilizes a strict pair-accuracy metric that requires a model to correctly answer questions for both original and counterfactual versions of a video.
+
+### Dataset Details
+DyBench consists of **3,014 videos** covering three primary categories of spatiotemporal dynamics:
+- **Reversible Dynamics**: Evaluating if models understand physical processes that can be temporally reversed.
+- **Moving Direction**: Tracking the spatial trajectory and direction of motion.
+- **Event Sequence**: Understanding the temporal order in which events occur.
+
+### Data Structure
+The dataset is organized into three configurations corresponding to the tasks above:
+- `event_sequence`
+- `moving_direction`
+- `reversible_dynamics`
+
+Each configuration contains JSON files mapping videos to their respective questions and ground-truth answers.