Commonsense Reasoning on CSQA2 (dev)
74AccuracyRetrieval-based knowledge
Evaluation results
| Method | Links | |
|---|---|---|
| Retrieval-based knowledgeInference Model=Unicorn-ft, Source=Google snippets2021.10 | 74 | |
| Generated Knowledge PromptingInference Model=Unicorn-ft2021.10 | 72.37 | |
| Context sentencesInference Model=Unicorn-ft2021.10 | 70.92 | |
| Random sentencesInference Model=Unicorn-ft2021.10 | 70.37 | |
| Vanilla baselineInference Model=Unicorn-ft2021.10 | 69.9 | |
| Unicorn-ft (prev. SOTA)2021.10 | 69.9 | |
| AnswersInference Model=Unicorn-ft2021.10 | 69.22 |