List your company

Fact Verification on FEVER (EM, $Δ$NF)

93.6Exact Match (EM)MADARA
86.8488.59590.3592.105Jun 23, 2026

Evaluation results

MethodLinks
MADARAModel=Gemma-2-9B, RSC=Quality-Ordered, ρ̂1=0.58, Strategy=ATF2026.0693.61.4
CoTModel=Mistral-7B-v0.32026.0692.6—
MADARAModel=Mistral-7B-v0.3, RSC=Quality-Ordered, ρ̂1=0.28, Strategy=CoT2026.0692.61.9
CoTModel=Gemma-2-9B2026.0692.5—
SDAModel=Gemma-2-9B2026.0692.5—
3-AgentModel=Gemma-2-9B2026.0692.4—
NFModel=Gemma-2-9B2026.0692.2—
MADARA (Average)2026.0691.21.5
MADARAModel=Llama-3.1-8B, RSC=Quality-Ordered, ρ̂1=0.65, Strategy=ATF2026.0690.92.2
3-AgentModel=Mistral-7B-v0.32026.0690.9—
MADARAModel=Qwen3-8B, RSC=Quality-Ordered, ρ̂1=0.63, Strategy=ATF2026.0690.91.3
NFModel=Mistral-7B-v0.32026.0690.7—
SDAModel=Mistral-7B-v0.32026.0690.7—
SDAModel=Qwen2.5-7B2026.0690.5—
AverageMethod_Group=SDA2026.0690.4—
AverageMethod_Group=CoT2026.0690.1—
SDAModel=Qwen3-8B2026.0690—
AverageMethod_Group=3-Agent2026.0690—
3-AgentModel=Qwen3-8B2026.0689.8—
AverageMethod_Group=NF2026.0689.7—
NFModel=Qwen3-8B2026.0689.6—
CoTModel=Qwen3-8B2026.0689.5—
3-AgentModel=Qwen2.5-7B2026.0688.9—
NFModel=Llama-3.1-8B2026.0688.7—
CoTModel=Llama-3.1-8B2026.0688.4—
SDAModel=Llama-3.1-8B2026.0688.2—
MADARAModel=Qwen2.5-7B, RSC=Quality-Ordered, ρ̂1=0.62, Strategy=ATF†2026.0688.21.1
3-AgentModel=Llama-3.1-8B2026.0687.8—
CoTModel=Qwen2.5-7B2026.0687.5—
NFModel=Qwen2.5-7B2026.0687.1—