List your company
ResearchFormerly WizWand

Commonsense Reasoning on SWAG (test)

0.9412AccuracyTEAM
0.7735520.8170760.86060.904124Apr 20, 2020Sep 20, 2020Feb 21, 2021Jul 25, 2021Dec 25, 2021May 28, 2022Oct 29, 2022
Updated 8mo ago

Evaluation results

MethodLinks
TEAMBackbone=DeBERTa Large2022.100.9412
UNICORN 11BModel=UNICORN 11B2022.100.939
XELECTRA-lModel Size=large2022.040.9283
ScoreBackbone=DeBERTa Large2022.100.9237
ALUMBackbone=ROBERTA-LARGE2020.040.91
ROBERTA_LARGE2020.040.899
TEAMBackbone=RoBERTa Large2022.100.8988
ScoreBackbone=RoBERTa Large2022.100.8897
XELECTRA-bModel Size=base2022.040.8891
Human2020.040.88
BERT_LARGE2020.040.863
XBERT-bModel Size=base2022.040.8134
GPT2020.040.78