Natural Language Inference On Anli Test
평가 지표
A1
A2
A3
평가 결과
이 벤치마크에서 각 모델의 성능 결과
비교 표
| 모델 이름 | A1 | A2 | A3 |
|---|---|---|---|
| a-systematic-study-and-comprehensive | 62.3 | 52.6 | 54.1 |
| language-models-are-few-shot-learners | 36.8 | 34 | 40.2 |
| palm-2-technical-report-1 | 53.1 | 48.8 | 53.2 |
| prompting-for-explanations-improves | 75.6 | 60.6 | 59.9 |
| knowledge-in-context-towards-knowledgeable | 36.30 | 35.00 | 37.60 |
| bloomberggpt-a-large-language-model-for | 33.6 | 33.8 | 35.17 |
| bloomberggpt-a-large-language-model-for | 33.1 | 34.2 | 34.92 |
| large-language-models-can-self-improve | - | 64.5 | 63.4 |
| large-language-models-can-self-improve | - | 66.5 | 67.9 |
| exploring-the-benefits-of-training-expert | 35.49 | 34.64 | 31.22 |
| bloomberggpt-a-large-language-model-for | 32.6 | 33.8 | 36.17 |
| prompting-for-explanations-improves | 81.8 | 72.5 | 74.8 |
| infobert-improving-robustness-of-language-1 | 75 | 50.5 | 47.7 |
| palm-2-technical-report-1 | 73.1 | 63.4 | 67.1 |
| the-cot-collection-improving-zero-shot-and | 41.7 | 37.2 | 41.9 |
| large-language-models-can-self-improve | - | 58.9 | 60.6 |
| roberta-a-robustly-optimized-bert-pretraining | 72.4 | 49.8 | 44.4 |
| large-language-models-can-self-improve | - | 64.8 | 66.9 |
| bloomberggpt-a-large-language-model-for | 32.9 | 34.4 | 37.33 |
| xlnet-generalized-autoregressive-pretraining | 70.3 | 50.9 | 49.4 |
| adversarial-training-for-large-neural | 72.3 | 52.1 | 48.4 |
| large-language-models-can-self-improve | - | 55.8 | 55.8 |
| palm-2-technical-report-1 | 58.1 | 49.5 | 54.5 |
| large-language-models-can-self-improve | - | 65.3 | 67.3 |
| guess-the-instruction-making-language-models | 39.99 | 37.05 | 37.73 |