Justification Quality Evaluation on AVeriTeC Retrieved (H) 50 correctly verified claims
3.67MOSDebateCV
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| DebateCVCriterion=Evidence Used, Evidence condition=Retrieved (H), Backbone=Llama-3.1-8B2025.07 | 3.67 | 57.6 | 27.5 | 14.9 | |
| DebateCVCriterion=Reasoning Pathway, Evidence condition=Retrieved (H), Backbone=Llama-3.1-8B2025.07 | 3.59 | 55.6 | 28.3 | 16.1 | |
| DebateCVCriterion=Sources of Uncertainty, Evidence condition=Retrieved (H), Backbone=Llama-3.1-8B2025.07 | 3.07 | 50.5 | 32.8 | 16.7 | |
| HerOCriterion=Evidence Used, Evidence condition=Retrieved (H)2025.07 | 2.6 | — | — | — | |
| HerOCriterion=Reasoning Pathway, Evidence condition=Retrieved (H)2025.07 | 2.49 | — | — | — | |
| HerOCriterion=Sources of Uncertainty, Evidence condition=Retrieved (H)2025.07 | 2.22 | — | — | — |