Explanation generation on Mocheg (test)
32.1ROUGE-1DePlot+FlanT5
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DePlot+FlanT5evidence_setting=gold evidence2026.02 | 32.1 | 15.5 | 14.7 | |
| MochegModelevidence_setting=gold evidence2026.02 | 31.4 | 15.4 | 14.4 | |
| MEVERevidence_setting=gold evidence, include_images=true2026.02 | 31.4 | 16.4 | 12.4 | |
| GPT-4oevidence_setting=gold evidence2026.02 | 31.2 | 10.1 | 4.7 | |
| JustiLMevidence_setting=gold evidence2026.02 | 30.7 | 15.8 | 12.9 | |
| MEVERevidence_setting=gold evidence, include_images=false2026.02 | 29.6 | 15.8 | 11.1 | |
| ECENetevidence_setting=gold evidence2026.02 | 29.2 | 15.2 | 12.4 |