Table 5 Summary of ChatGPT (model 3.5) to generate responses with less than 10% matching
From: Evaluating the authenticity of ChatGPT responses: a study on text-matching capabilities
Source of Matching | Observed Capability | Expected Capability |
|---|---|---|
Overall | 33.33% | 66.5% |
ChatGPT previously generated responses | 50% | 77.6% |
students' global database | 86.7% | 97.3% |
Internet | 80% | 94% |