Claim Verification on AVeriTeC Retrieved (H) (dev)
72.8AccuracyDebateCV
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DebateCVLLMs=Llama-3.1-8B2025.07 | 72.8 | 54.4 | |
| DebateCVLLMs=Qwen-2.5-7B2025.07 | 72.6 | 54.6 | |
| w/o CcorrectLLMs=Llama-3.1-8B2025.07 | 70.8 | 53.8 | |
| HerOLLMs=Llama-3.1-8B2025.07 | 70.2 | 52.6 | |
| w/o CcorrectLLMs=Qwen-2.5-7B2025.07 | 69.4 | 53 | |
| InFactLLMs=GPT-4o2025.07 | 68.6 | 52.6 | |
| HerOLLMs=Qwen-2.5-7B2025.07 | 68.2 | 50.6 | |
| CooperationLLMs=GPT-4o2025.07 | 67.6 | 51.2 | |
| w/o Debate-SFTLLMs=GPT-4o2025.07 | 65.4 | 48.6 | |
| InFactLLMs=GPT-4o-Mini2025.07 | 65.2 | 47.8 | |
| CooperationLLMs=GPT-4o-Mini2025.07 | 65.2 | 49.4 | |
| MajorityLLMs=GPT-4o2025.07 | 64.4 | 49.2 | |
| CoTLLMs=GPT-4o2025.07 | 63.8 | 47.8 | |
| MajorityLLMs=GPT-4o-Mini2025.07 | 63.8 | 47.8 | |
| MajorityLLMs=Llama-3.1-8B2025.07 | 63.8 | 47.6 | |
| InFactLLMs=Qwen-2.5-7B2025.07 | 63.6 | 47.4 | |
| MajorityLLMs=Qwen-2.5-7B2025.07 | 63.6 | 48.2 | |
| InFactLLMs=Llama-3.1-8B2025.07 | 63.4 | 47 | |
| w/o CerrorLLMs=Llama-3.1-8B2025.07 | 63 | 49.2 | |
| w/o CerrorLLMs=Qwen-2.5-7B2025.07 | 62.9 | 49.1 | |
| CoTLLMs=GPT-4o-Mini2025.07 | 62.8 | 46.8 | |
| CoTLLMs=Qwen-2.5-7B2025.07 | 62.6 | 46.8 | |
| CoTLLMs=Llama-3.1-8B2025.07 | 62.4 | 46.6 | |
| CooperationLLMs=Llama-3.1-8B2025.07 | 61.2 | 46.8 | |
| CooperationLLMs=Qwen-2.5-7B2025.07 | 60.8 | 46.4 | |
| w/o Debate-SFTLLMs=GPT-4o-Mini2025.07 | 54.8 | 40.4 | |
| w/o Debate-SFTLLMs=Qwen-2.5-7B2025.07 | 54.2 | 41 | |
| w/o Debate-SFTLLMs=Llama-3.1-8B2025.07 | 53.8 | 41.2 |