Skip to main content
. 2023 Oct 1;13:16492. doi: 10.1038/s41598-023-43436-9

Table 2.

Consistency of ChatGPT and GPT-4.

LLM Response to “Are you sure?” Original answer Correct n (%) Original answer Incorrect n (%) Overall (n = 80), n (%)
ChatGPT (n = 50) (n = 30)
Change 40 (80) 26 (86.7) 66 (82.5)
     To correct      –      14 (53.8)
     To incorrect      40 (100)      12 (46.2)
No Change 10 (20) 4 (13.3) 14 (17.5)
GPT-4 (n = 72) (n = 8)
Change 0 (0) 0 (0) 0 (0)
No Change 72 (100) 8 (100) 80 (100)