List your company

Question Answering on 2WikiMultiHopQA out-domain (test)

56.32LasJD2PLAN-7B
29.477636.446343.41550.3837Jan 13, 2026

Evaluation results

MethodLinks
D2PLAN-7BBase LLM=Qwen2.5-7B-Instruct2026.0156.3250.02
D2PLAN-3BBase LLM=Qwen2.5-3B-Instruct2026.0152.5345.9
R1-SearcherBase LLM=Qwen2.5-7B-Instruct2026.0152.0446.13
StepSearchBase LLM=Qwen2.5-7B-Instruct2026.0149.5742.89
AutoRefineBase LLM=Qwen2.5-3B-Instruct2026.0148.0838.05
ReSearchBase LLM=Qwen2.5-7B-Instruct2026.0147.3441.01
Search-R1Base LLM=Qwen2.5-7B-Instruct2026.0139.3633.13
ZeroSearchBase LLM=Qwen2.5-7B-Instruct2026.0136.731.76
StepSearchBase LLM=Qwen2.5-3B-Instruct2026.0133.3328.78
Search-R1Base LLM=Qwen2.5-3B-Instruct2026.0133.0620.92
ZeroSearchBase LLM=Qwen2.5-3B-Instruct2026.0130.5126.42