Fig. 9
From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilities
The capability of ChatGPT (model 4) to generate responses with less than 10% matching (Overall)
Fig. 9 | International Journal for Educational Integrity