dataset
active
dataset:hellaswag

HellaSwag

Commonsense reasoning benchmark used to test whether LLM-vision alignment predicts downstream performance

Neighborhood — ranked by edge-count