Fig. 8
From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilities
The capability of ChatGPT (model 3.5) to generate responses with less than 25% matching (Internet)
Fig. 8 | International Journal for Educational Integrity