Explanation Generation on ChartCheck (test)
48.9ROUGE-1MEVER
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MEVERevidence_setting=gold evidence, include_images=true2026.02 | 48.9 | 27.4 | 16.8 | |
| ChartGemmaevidence_setting=gold evidence2026.02 | 48 | 26.6 | 16.6 | |
| UniChartevidence_setting=gold evidence2026.02 | 47.6 | 26.5 | 15.6 | |
| MEVERevidence_setting=gold evidence, include_images=false2026.02 | 47.6 | 26.5 | 16.7 | |
| ECENetevidence_setting=gold evidence2026.02 | 47.4 | 26.4 | 15.2 | |
| DePlot+FlanT5evidence_setting=gold evidence2026.02 | 45.4 | 25.6 | 15.4 | |
| MochegModelevidence_setting=gold evidence2026.02 | 45.3 | 25.4 | 15.6 | |
| JustiLMevidence_setting=gold evidence2026.02 | 40 | 22.5 | 14.4 | |
| GPT-4oevidence_setting=gold evidence2026.02 | 23 | 9.6 | 13.2 |