Reading Comprehension On Reclor
評価指標
Test
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | Test | Paper Title | Repository |
|---|---|---|---|
| WWZ | 69.7 | - | - |
| 5BitClan | 56.9 | - | - |
| BERT-large | 49.8 | - | - |
| DeBERTa-v2-xxlarge-AMR-LE-Contraposition | 77.2 | - | - |
| TwistedFate | 25.3 | - | - |
| Gariscat | 40.5 | - | - |
| MERIT w/ reasoning aware | 63.2 | - | - |
| BERT-large+MNLI | 50.3 | - | - |
| alsace | 59.2 | - | - |
| MERIt-deberta-v2-xxlarge deberta.v2.xxlarge.path.override_True.norm_1.1.0.w2.A100.cp200.s42 | 79.3 | - | - |
| Knowledge model | 79.2 | - | - |
| RoBERTa-single | 63.5 | Logiformer: A Two-Branch Graph Transformer Network for Interpretable Logical Reasoning | |
| ALBERT-XXLarge-V2 | 62.6 | - | - |
| xlnet-large-uncased [extended data] | 69.3 | - | - |
| RoBERTa-single | 58.9 | Fact-driven Logical Reasoning for Machine Reading Comprehension | |
| XLNet-base | 50.4 | ReClor: A Reading Comprehension Dataset Requiring Logical Reasoning | |
| BachTE | 26.0 | - | - |
| ro_DA_CE_v100_224 | 61.7 | - | - |
| ELECTRA and ALBERT | 71.0 | Answer Uncertainty and Unanswerability in Multiple-Choice Machine Reading Comprehension | - |
| Tournament2 | 66.7 | - | - |
0 of 39 row(s) selected.