Fig. 10From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilitiesThe capability of ChatGPT (model 4) to generate responses with less than25% matching (Overall)Back to article page