Replace with Croissant-friendly questions table

README.md

@@ -10,12 +10,6 @@ configs:
 
 BEHAVIOR ESI-Bench is a question dataset for evaluating embodied spatial intelligence across indoor scenes, object arrangements, physical reasoning, temporal understanding, and active exploration tasks.
 
-The Hugging Face release is centered on a single fixed-column JSONL table:
-
-```text
-data/questions.jsonl
-```
-
 Each row is one question instance. The table is intentionally flattened for Hugging Face Dataset Viewer and Croissant compatibility.
 
 ## Schema
@@ -56,12 +50,6 @@ Temporal Understanding
 
 The `small_task` column stores the corresponding subtask, and `runner_task` stores the internal task module name used by the original BEHAVIOR active-exploration code.
 
-## Notes
-
-Some questions reference rendered image filenames in `image_paths_json`. These are clean relative asset names rather than machine-specific absolute paths.
-
-Questions with unavailable source metadata are retained as explicit missing placeholders, with `question` and `answer` set to `null`.
-
 ## Croissant
 
 Hugging Face automatically generates Croissant metadata from the Dataset Viewer once this dataset is processed. The Croissant JSON-LD endpoint is:
@@ -69,11 +57,3 @@ Hugging Face automatically generates Croissant metadata from the Dataset Viewer
 ```text
 https://huggingface.co/api/datasets/ESI-Bench/esi-bench/croissant
 ```
-
-## Rebuild
-
-The table can be regenerated from the internal task JSON tree with:
-
-```bash
-python scripts/export_hf_questions.py
-```