Common Sense Reasoning On Codah

Accuracy

Results

Performance results of various models on this benchmark

Comparison Table

Model Name	Accuracy
aqua-an-adversarially-authored-question	69.6