AHA dataset

anchor

Note

introduced by the paper that uses it, so the anchor is that paper; the dataset has no separate release of its own that the source prints

Findings

• IC-005GPT-4o underperforms AHA and other VLMs in detecting and reasoning about robotic manipulation failures across multiple datasets. [eval]