Commonsense Reasoning on NumerSense core (test)
79.24AccuracyGenerated Knowledge Prompting
Evaluation results
| Method | Links | |
|---|---|---|
| Generated Knowledge PromptingInference Model=T5-11b2021.10 | 79.24 | |
| T5-11b 1.1 +digitsNote=Previous SOTA2021.10 | 72.61 | |
| Retrieval-based knowledgeInference Model=T5-11b2021.10 | 70.41 | |
| Vanilla baselineInference Model=T5-11b2021.10 | 70.23 |