List your company

General Question Answering on PopQA Pop. (OSR%)

27.7OSR ScoreStepSearch-base
-1.1086.37113.8521.329Apr 19, 2026

Evaluation results

MethodLinks
StepSearch-baseBackbone=Qwen2.5-3B2026.0427.7
StepSearch-instructBackbone=Qwen2.5-7B2026.0426.6
StepSearch-baseBackbone=Qwen2.5-7B2026.0412.1
Search-R1-baseBackbone=Qwen2.5-7B2026.0411.4
Search-R1-instructBackbone=Qwen2.5-3B2026.048.9
HIPRAG-instructBackbone=Qwen2.5-7B2026.047.9
Search-R1-instructBackbone=Qwen2.5-7B2026.046.8
Search-R1-baseBackbone=Qwen2.5-3B2026.045.6
HIPRAG-baseBackbone=Qwen2.5-3B2026.045.5
StepSearch-instructBackbone=Qwen2.5-3B2026.044.9
HIPRAG-instructBackbone=Qwen2.5-3B2026.041.8
AutoSearchBackbone=Qwen2.5-7B2026.040.5
AutoSearchBackbone=Qwen2.5-3B2026.040