Improve metadata: Update task category and add relevant tags

This PR updates the `task_categories` metadata from `image-text-to-text` to `image-feature-extraction`, as the dataset is primarily used for diagnosing visual feature extraction and compositional failures in Vision-Language Models (VLMs), as described in the paper.

Additionally, relevant tags such as `vlm`, `mechanistic-interpretability`, `compositional-reasoning`, and `synthetic-data` have been added to enhance the dataset's discoverability and better reflect its characteristics and research focus.

README.md

@@ -1,56 +1,62 @@
----
-dataset_info:
-  features:
-  - name: num_shapes
-    dtype: int64
-  - name: shape_0
-    dtype: string
-  - name: color_0
-    dtype: string
-  - name: position_0
-    dtype: string
-  - name: shape_1
-    dtype: string
-  - name: color_1
-    dtype: string
-  - name: position_1
-    dtype: string
-  - name: shape_2
-    dtype: string
-  - name: color_2
-    dtype: string
-  - name: position_2
-    dtype: string
-  - name: shape_3
-    dtype: string
-  - name: color_3
-    dtype: string
-  - name: position_3
-    dtype: string
-  - name: shape_4
-    dtype: string
-  - name: color_4
-    dtype: string
-  - name: position_4
-    dtype: string
-  - name: image
-    dtype: image
-  splits:
-  - name: train
-    num_bytes: 1559700
-    num_examples: 500
-  download_size: 1354975
-  dataset_size: 1559700
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-pretty_name: Toy Shapes Dataset
-arxiv: 2508.16652
-task_categories:
-- image-text-to-text
----
+---
+task_categories:
+- image-feature-extraction
+tags:
+- vlm
+- mechanistic-interpretability
+- compositional-reasoning
+- synthetic-data
+pretty_name: Toy Shapes Dataset
+dataset_info:
+  features:
+  - name: num_shapes
+    dtype: int64
+  - name: shape_0
+    dtype: string
+  - name: color_0
+    dtype: string
+  - name: position_0
+    dtype: string
+  - name: shape_1
+    dtype: string
+  - name: color_1
+    dtype: string
+  - name: position_1
+    dtype: string
+  - name: shape_2
+    dtype: string
+  - name: color_2
+    dtype: string
+  - name: position_2
+    dtype: string
+  - name: shape_3
+    dtype: string
+  - name: color_3
+    dtype: string
+  - name: position_3
+    dtype: string
+  - name: shape_4
+    dtype: string
+  - name: color_4
+    dtype: string
+  - name: position_4
+    dtype: string
+  - name: image
+    dtype: image
+  splits:
+  - name: train
+    num_bytes: 1559700
+    num_examples: 500
+  download_size: 1354975
+  dataset_size: 1559700
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+arxiv: 2508.16652
+---
+
 This dataset is used in the work **"Do VLMs Have Bad Eyes? Diagnosing Compositional Failures via Mechanistic Interpretability"** [Paper](https://huggingface.co/papers/2508.16652)
 
 For more details, please refer to the [GitHub repository](https://github.com/Mystic-Slice/Do-VLMs-Have-Bad-Eyes).