List your company

Question Answering on PopQA Normal (test)

25.95F1 ScoreOpenDecoder
6.60611.62816.6521.672Jan 13, 2026

Evaluation results

MethodLinks
OpenDecoderBackbone=Qwen-2.5-3B-Instruct2026.0125.9522.8
RbFTBackbone=Qwen-2.5-3B-Instruct2026.0124.7321.42
AstuteRAGBackbone=Qwen-2.5-3B-Instruct2026.0123.9219.9
Vanilla SFTBackbone=Qwen-2.5-3B-Instruct2026.0120.4617.37
InstructRAGBackbone=Qwen-2.5-3B-Instruct2026.0116.1921.7
No RAGBackbone=Qwen-2.5-3B-Instruct2026.0111.07—
RobustRAGBackbone=Qwen-2.5-3B-Instruct2026.0110.9115.9
Vanilla RAGBackbone=Qwen-2.5-3B-Instruct2026.017.3521.87