Explanation generation on AIChartClaim (test)
42.9ROUGE-1MEVER
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| MEVERevidence_setting=gold evidence, include_images=true2026.02 | 42.9 | 21.9 | 11.6 | |
| ChartGemmaevidence_setting=gold evidence2026.02 | 42.2 | 21.3 | 10.4 | |
| MEVERevidence_setting=gold evidence, include_images=false2026.02 | 41.8 | 20.8 | 10.6 | |
| UniChartevidence_setting=gold evidence2026.02 | 41.6 | 20.9 | 11 | |
| ECENetevidence_setting=gold evidence2026.02 | 41.5 | 20.4 | 10.2 | |
| DePlot+FlanT5evidence_setting=gold evidence2026.02 | 39.7 | 19.4 | 8.9 | |
| MochegModelevidence_setting=gold evidence2026.02 | 39.5 | 19.3 | 8.9 | |
| JustiLMevidence_setting=gold evidence2026.02 | 31.4 | 14 | 7 | |
| GPT-4oevidence_setting=gold evidence2026.02 | 25.5 | 18.1 | 6.8 |