Table 8 Results for on-the-fly findings to medical terms prediction using the generative LLM ChatGPT
From: Explanatory argumentation in natural language for correct and incorrect medical diagnoses
Prompting Method | Accuracy |
|---|---|
IO ChatGPT 4 | 0.64 |
IO ChatGPT 3 | 0.52 |
CoT ChatGPT 4 | 0.66 |
CoT ChatGPT 3 | 0.52 |
SC IO ChatGPT 4 | 0.64 |
SC IO ChatGPT 3 | 0.54 |
SC CoT ChatGPT 4 | 0.66 |
SC CoT ChatGPT 3 | 0.54 |