List your company

Long-form Question Answering on Biography FullP (100% polluted)

61.87VeriScore F1@kMIRAGE
10.275623.670337.06550.4597Jul 6, 2026

Evaluation results

MethodLinks
MIRAGEBackbone=GPT-4o-mini2026.0761.87
No-RAGBackbone=GPT-4o-mini2026.0758.99
MIRAGEBackbone=Qwen3-8B2026.0755.49
No-RAGBackbone=Qwen3-8B2026.0753.75
AstuteRAGBackbone=GPT-4o-mini2026.0747.91
RetRobust-NLIBackbone=GPT-4o-mini2026.0745.33
BRIDGEBackbone=GPT-4o-mini2026.0743.5
SelfRAGBackbone=GPT-4o-mini2026.0738.55
FLAREBackbone=GPT-4o-mini2026.0737.38
DRAGINBackbone=GPT-4o-mini2026.0732.09
MADAMRAGBackbone=GPT-4o-mini2026.0725.48
RobustRAGBackbone=Qwen3-8B2026.0725.22
TrustRAGBackbone=GPT-4o-mini2026.0723.42
RARGBackbone=GPT-4o-mini2026.0722.92
RAGBackbone=GPT-4o-mini2026.0722.65
BRIDGEBackbone=Qwen3-8B2026.0721.31
AstuteRAGBackbone=Qwen3-8B2026.0720.7
RobustRAGBackbone=GPT-4o-mini2026.0720.27
SelfRAGBackbone=Qwen3-8B2026.0718.97
DRAGINBackbone=Qwen3-8B2026.0717.5
RARGBackbone=Qwen3-8B2026.0716.07
TrustRAGBackbone=Qwen3-8B2026.0715.68
MADAMRAGBackbone=Qwen3-8B2026.0715.15
RAGBackbone=Qwen3-8B2026.0713.24
RetRobust-NLIBackbone=Qwen3-8B2026.0712.37
FLAREBackbone=Qwen3-8B2026.0712.26