List your company
ResearchFormerly WizWand

Multi-hop Question Answering on MuSiQue standard (val)

22.6Exact Match (EM)StepSearch-instruct
-0.6965.35211.417.448Mar 14, 2026
Updated 7mo ago

Evaluation results

MethodLinks
StepSearch-instructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.0322.6
ZeroSearch-instructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.0318.4
StepSearch-instructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.0317.4
Apex-SearcherType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.0316.4
Search-R1-InstructType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.0314.6
Apex-SearcherType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.0313.6
ReasonRAGType=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.0312.8
ZeroSearch-instructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.039.8
IRCoTType=Iterative RAG, Backbone=Qwen2.5-7B-Instruct2026.037.2
IRCoTType=Iterative RAG, Backbone=Qwen2.5-3B-Instruct2026.036.7
RAGType=Standard RAG, Backbone=Qwen2.5-7B-Instruct2026.035.8
Search-o1Type=Agentic RAG, Backbone=Qwen2.5-7B-Instruct2026.035.8
Search-o1Type=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.035.4
Search-R1-InstructType=Agentic RAG, Backbone=Qwen2.5-3B-Instruct2026.034.9
RAGType=Standard RAG, Backbone=Qwen2.5-3B-Instruct2026.034.7
Direct InferenceType=Non-Retrieval, Backbone=Qwen2.5-7B-Instruct2026.033.1
CoTType=Non-Retrieval, Backbone=Qwen2.5-7B-Instruct2026.032.2
Direct InferenceType=Non-Retrieval, Backbone=Qwen2.5-3B-Instruct2026.032
CoTType=Non-Retrieval, Backbone=Qwen2.5-3B-Instruct2026.030.2