How Reasonable are Common-Sense Reasoning Tasks: A Case-Study on the Winograd Schema Challenge and SWAG

Open paper

Benchmark	Model	Rank	Results
coreference-resolution-on-winograd-schema	GPT-2 Medium 774M (partial scoring)	#31	Accuracy: 69.2
coreference-resolution-on-winograd-schema	GPT-2 Medium 774M (full scoring)	#37	Accuracy: 64.5
coreference-resolution-on-winograd-schema	GPT-2 Small 117M (partial scoring)	#45	Accuracy: 61.5
coreference-resolution-on-winograd-schema	GPT-2 Small 117M (full scoring)	#54	Accuracy: 55.7