Question Answering On Copa
评估指标
Accuracy
评测结果
各个模型在此基准测试上的表现结果
比较表格
| 模型名称 | Accuracy |
|---|---|
| winogrande-an-adversarial-winograd-schema | 90.6 |
| efficient-language-modeling-with-sparse-all | 64 |
| finetuned-language-models-are-zero-shot | 94 |
| winogrande-an-adversarial-winograd-schema | 86.4 |
| finetuned-language-models-are-zero-shot | 91 |
| designing-effective-sparse-expert-models | 91 |
| exploring-the-limits-of-transfer-learning | 92 |
| language-models-are-few-shot-learners | 92 |
| hungry-hungry-hippos-towards-language | 67 |
| toward-efficient-language-model-pretraining | 99.4 |
| hungry-hungry-hippos-towards-language | 51 |
| exploring-the-limits-of-transfer-learning | 94.8 |
| the-cot-collection-improving-zero-shot-and | 90.9 |
| handling-multiword-expressions-in-causality | 69.9 |
| ask-me-anything-a-simple-strategy-for | 77.0 |
| knowledge-in-context-towards-knowledgeable | 85.30 |
| toward-efficient-language-model-pretraining | 98.2 |
| designing-effective-sparse-expert-models | 99.2 |
| socialiqa-commonsense-reasoning-about-social | 80.8 |
| palm-2-technical-report-1 | 90.0 |
| alexatm-20b-few-shot-learning-using-a-large | 78.0 |
| kelm-knowledge-enhanced-pre-trained-language | 78.0 |
| bloomberggpt-a-large-language-model-for | 84 |
| language-models-are-few-shot-learners | 73.0 |
| palm-2-technical-report-1 | 89.0 |
| handling-multiword-expressions-in-causality | 71.2 |
| unifying-language-learning-paradigms | 85 |
| palm-scaling-language-modeling-with-pathways-1 | 100 |
| unifying-language-learning-paradigms | 99 |
| handling-multiword-expressions-in-causality | 70.2 |
| hungry-hungry-hippos-towards-language | 67 |
| ask-me-anything-a-simple-strategy-for | 58.2 |
| efficient-language-modeling-with-sparse-all | 75 |
| guess-the-instruction-making-language-models | 89.88 |
| exploring-the-limits-of-transfer-learning | 83.4 |
| language-models-are-few-shot-learners | 86 |
| finetuned-language-models-are-zero-shot | 87 |
| deberta-decoding-enhanced-bert-with | 98.4 |
| socialiqa-commonsense-reasoning-about-social | 83.4 |
| handling-multiword-expressions-in-causality | 58.8 |
| efficient-language-modeling-with-sparse-all | 63 |
| bloomberggpt-a-large-language-model-for | 86 |
| hungry-hungry-hippos-towards-language | 77 |
| winogrande-an-adversarial-winograd-schema | 76.4 |
| bloomberggpt-a-large-language-model-for | 88 |
| ask-me-anything-a-simple-strategy-for | 84.0 |
| winogrande-an-adversarial-winograd-schema | 65.4 |
| exploring-the-benefits-of-training-expert | 79.25 |
| palm-2-technical-report-1 | 96.0 |
| language-models-are-few-shot-learners | 91 |
| winogrande-an-adversarial-winograd-schema | 84.4 |
| deberta-decoding-enhanced-bert-with | 96.8 |
| bloomberggpt-a-large-language-model-for | 86 |
| hungry-hungry-hippos-towards-language | 81 |
| n-grammer-augmenting-transformers-with-latent-1 | 60.0 |
| efficient-language-modeling-with-sparse-all | 76 |
| exploring-the-limits-of-transfer-learning | 71.2 |
| language-models-are-few-shot-learners | 87 |
| efficient-language-modeling-with-sparse-all | 79 |
| handling-multiword-expressions-in-causality | 50 |