List your company

Logical Reasoning on LogiQA 1.0 (test)

86AccuracyHuman Performance
22.5639.0355.571.97May 2, 2022Jun 9, 2022Jul 18, 2022Aug 25, 2022Oct 3, 2022Nov 10, 2022Dec 19, 2022

Evaluation results

MethodLinks
Human Performance2022.0586
APOLLOBase Architecture=DeBERTa-v2-xxlarge2022.1251
DeBERTa-v2-xxlargeBase Architecture=DeBERTa-v2-xxlarge2022.1249.8
APOLLOBase Architecture=DeBERTa-v32022.1244.4
LogiformerEncoder=RoBERTa-Large2022.0542.55
FocalReasonerEncoder=RoBERTa-Large2022.0540.25
DeBERTa-v3Base Architecture=DeBERTa-v32022.1240.1
DAGNEncoder=RoBERTa-Large2022.0539.32
RoBERTa-LargeEncoder=RoBERTa-Large2022.0535.33
BERT-LargeEncoder=BERT-Large2022.0531.03
Random2022.0525