Table 6 Summary of ChatGPT (model 3.5) to generate responses with less than 25% matching
From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilities
Source of Matching | Observed Capability | Expected Capability |
|---|---|---|
Overall | 73.3% | 76.6% |
ChatGPT previously generated responses | 83.3% | 84.4% |
students' global database | 93.3% | 98.8% |
Internet | 96.7% | 96.8% |