List your company

Multi-hop QA on Bamboogle (EM, F1)

56EMSearch-o1
1.08815.34429.643.856Jan 9, 2025Mar 20, 2025May 30, 2025Aug 9, 2025Oct 19, 2025Dec 29, 2025Mar 10, 2026

Evaluation results

MethodLinks
Search-o1Reasoning Protocol=Retrieval-augmented Reasoning with Reason-in-Documents, Parameter Count=32B2025.015667.8
RAG-QwQ-32BReasoning Protocol=Retrieval-augmented Reasoning, Parameter Count=32B2025.0155.267.4
Llama3.3-70BReasoning Protocol=Direct Reasoning (w/o Retrieval), Parameter Count=70B2025.0154.467.8
RAgent-Qwen2.5-32BReasoning Protocol=Retrieval-augmented Reasoning, Parameter Count=32B2025.0154.466.4
RAG-Qwen2.5-32BReasoning Protocol=Retrieval-augmented Reasoning, Parameter Count=32B2025.015266
RAgent-QwQ-32BReasoning Protocol=Retrieval-augmented Reasoning, Parameter Count=32B2025.015264.7
Qwen2.5-32BReasoning Protocol=Direct Reasoning (w/o Retrieval), Parameter Count=32B2025.0149.663.2
Qwen2.5-72BReasoning Protocol=Direct Reasoning (w/o Retrieval), Parameter Count=72B2025.0147.261.7
TaSR-RAGBackbone=Qwen2.5-72B-Instruct2026.0345.649.1
StructRAGBackbone=Qwen2.5-72B-Instruct2026.0344.854.5
QwQ-32BReasoning Protocol=Direct Reasoning (w/o Retrieval), Parameter Count=32B2025.0138.453.7
HippoRAGBackbone=Qwen2.5-72B-Instruct2026.0329.635.9
TaSR-RAGBackbone=Qwen2.5-7B-Instruct2026.0325.628.6
IRCoTBackbone=Qwen2.5-72B-Instruct2026.0323.236.6
CoTBackbone=Qwen2.5-7B-Instruct2026.0323.220.8
IRCoTBackbone=Qwen2.5-7B-Instruct2026.0322.432.1
RAGBackbone=Qwen2.5-72B-Instruct2026.0321.634.1
RAGBackbone=Qwen2.5-7B-Instruct2026.0320.822.4
CoTBackbone=Qwen2.5-72B-Instruct2026.0319.631
HyperGraphRAGBackbone=Qwen2.5-72B-Instruct2026.0319.227
Direct InferenceBackbone=Qwen2.5-72B-Instruct2026.0317.627.8
HippoRAGBackbone=Qwen2.5-7B-Instruct2026.0315.223.7
StructRAGBackbone=Qwen2.5-7B-Instruct2026.0313.624.5
Direct InferenceBackbone=Qwen2.5-7B-Instruct2026.031217.2
GraphRAGBackbone=Qwen2.5-7B-Instruct2026.03814.9
GraphRAGBackbone=Qwen2.5-72B-Instruct2026.036.415.8
HypergraphRAGBackbone=Qwen2.5-7B-Instruct2026.033.23.9