Factual Inconsistency Detection on CHOCOLATE LLM
0.205Kendall's TauGPT-4V
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4V2023.12 | 0.205 | |
| DePlot + GPT-42023.12 | 0.117 | |
| Bard2023.12 | 0.105 | |
| CHARTVE2023.12 | 0.091 | |
| LLaVA-1.5-13B2023.12 | 0.057 | |
| ChartLlama2023.12 | 0.057 | |
| ChartAssistant-S2023.12 | 0.057 | |
| QAFACTEVAL2023.12 | 0.045 | |
| SUMMAC2023.12 | 0.023 |