List your company

Single-Hop Question Answering on NQ (Accuracy)

49.1AccuracyMultiSearch
0.42813.06425.738.336May 13, 2026

Evaluation results

MethodLinks
MultiSearchBackbone=Qwen2.5-7B, Model Variant=Base2026.0549.1
MultiSearchBackbone=Qwen2.5-7B, Model Variant=Instruct2026.0547.2
MultiSearchBackbone=Qwen2.5-3B, Model Variant=Base2026.0547.1
MultiSearchBackbone=Qwen2.5-3B, Model Variant=Instruct2026.0547
AutoRefineBackbone=Qwen2.5-7B, Model Variant=Base2026.0547
AutoRefineBackbone=Qwen2.5-3B, Model Variant=Base2026.0546.7
AutoRefineBackbone=Qwen2.5-3B, Model Variant=Instruct2026.0543.6
ZeroSearchBackbone=Qwen2.5-3B, Model Variant=Base2026.0543.4
ZeroSearchBackbone=Qwen2.5-7B, Model Variant=Base2026.0543
Search-R1Backbone=Qwen2.5-7B, Model Variant=Instruct2026.0542.9
ReSearchBackbone=Qwen2.5-3B, Model Variant=Base2026.0542.7
Search-R1Backbone=Qwen2.5-3B, Model Variant=Base2026.0542.1
AutoRefineBackbone=Qwen2.5-7B, Model Variant=Instruct2026.0541.9
ZeroSearchBackbone=Qwen2.5-7B, Model Variant=Instruct2026.0541.4
Dr.ZeroBackbone=Qwen2.5-7B2026.0540.6
ZeroSearchBackbone=Qwen2.5-3B, Model Variant=Instruct2026.0540.2
Search-R1Backbone=Qwen2.5-3B, Model Variant=Instruct2026.0539.7
Search-R1Backbone=Qwen2.5-7B, Model Variant=Base2026.0539.5
Dr.ZeroBackbone=Qwen2.5-3B2026.0539.1
AdaSearchBackbone=Qwen2.5-3B2026.0537.9
ReSearchBackbone=Qwen2.5-3B, Model Variant=Instruct2026.0536.5
RAGBackbone=Qwen2.5-7B2026.0534.9
RAGBackbone=Qwen2.5-3B2026.0534.8
SFTBackbone=Qwen2.5-7B2026.0531.8
R1Backbone=Qwen2.5-7B, Model Variant=Base2026.0529.7
R1Backbone=Qwen2.5-7B, Model Variant=Instruct2026.0527
SFTBackbone=Qwen2.5-3B2026.0524.9
Search-o1Backbone=Qwen2.5-3B2026.0523.8
R1Backbone=Qwen2.5-3B, Model Variant=Base2026.0522.6
IRCoTBackbone=Qwen2.5-7B2026.0522.4
R1Backbone=Qwen2.5-3B, Model Variant=Instruct2026.0521
Search-o1Backbone=Qwen2.5-7B2026.0515.1
Direct InferenceBackbone=Qwen2.5-7B2026.0513.4
IRCoTBackbone=Qwen2.5-3B2026.0511.1
Direct GenerationBackbone=Qwen2.5-3B2026.0510.6
CoTBackbone=Qwen2.5-7B2026.054.8
CoTBackbone=Qwen2.5-3B2026.052.3