Simplify dataset columns and label IID/OOD only for Drag

README.md

@@ -71,9 +71,10 @@ giving 450 scene pairs and 900 episodes. The `pair_id` identifies the shared
 scene; the `episode_id` identifies a particular environment episode.
 
 All Hugging Face configurations expose a `test` split. The default `all`
-configuration combines the same six subsets into 900 rows. `distribution`
-preserves each episode's original IID/OOD designation; each Drag environment
-contains 75 IID and 75 OOD episodes.
+configuration combines the same six subsets into 900 rows. The `distribution`
+field identifies the 75 IID and 75 OOD episodes in each Drag environment.
+Rotation and Ten-Choice each use fixed test sets without an IID/OOD subdivision,
+so their `distribution` values are `null`.
 
 ## Load episode records
 
@@ -92,11 +93,10 @@ episode = json.loads(drag[0]["episode_json"])
 | `episode_id`, `pair_id` | Episode and paired-scene identifiers |
 | `suite_id`, `variant`, `family` | Benchmark suite, environment, and task family |
 | `instruction` | Task instruction |
-| `exploration_level`, `distribution` | Episode categories |
-| `case_seed` | Scene seed |
-| `viewport_width`, `viewport_height` | Viewport size in pixels |
+| `exploration_level` | Exploration category |
+| `distribution` | `iid` or `ood` for Drag; `null` for Rotation and Ten-Choice |
 | `canonical_case_path` | Scene metadata path relative to the dataset root |
-| `episode_json` | Complete original episode configuration serialized as JSON |
+| `episode_json` | Complete runtime episode configuration serialized as JSON |
 
 The runtime consumes the full configuration, including hidden dynamics and
 success criteria. Agent observations consist of task instructions, screenshots,