Table 1 Performance metrics of the GPT-3 model in the zero-shot and few-shot setting
From: Leveraging pre-trained language models for mining microbiome-disease relationships
Model | Accuracy | F1 score | Precision | Recall |
|---|---|---|---|---|
GPT-3 (zero-shot) | 0.48 | 0.50 | 0.58 | 0.48 |
GPT-3 (few-shot) | 0.57 | 0.56 | 0.57 | 0.57 |