Commonsense Reasoning on Winograd Schema Challenge (WSC) (test)
75.1AccuracyHNN
Evaluation results
| Method | Links | |
|---|---|---|
| HNN2019.07 | 75.1 | |
| BERTWiki-WSCR2019.07 | 72.2 | |
| GPT-22019.07 | 70.7 | |
| BERTWSCR2019.07 | 70.3 | |
| DSSM2019.07 | 63 | |
| UDSSM-IIensemble=true2019.04 | 62.4 | |
| BERTLARGE-LM2019.07 | 62 | |
| Google Language Modelensemble=true2019.04 | 61.5 | |
| UDSSM-IIensemble=false2019.04 | 59.2 | |
| UDSSM-Iensemble=true2019.04 | 57.1 | |
| Google Language Modelensemble=false2019.04 | 56.4 | |
| UDSSM-Iensemble=false2019.04 | 54.5 | |
| USSM + Cause-Effect + WordNet + ConceptNet Knowledge BasesKnowledge Bases=Cause-Effect, WordNet, ConceptNet2019.04 | 52.8 | |
| USSM + NKAM + 3 Knowledge BasesKnowledge Bases=32019.04 | 52.8 | |
| USSM + Cause-Effect Knowledge BaseKnowledge Base=Cause-Effect2019.04 | 52 | |
| ELMo2019.04 | 51.5 | |
| Co-reference Resolution Tool2019.04 | 50.5 |