List your company

Multi-Hop QA on FictionalHot

6.1Exact Match (EM)ReSeek
-0.141.483.14.72Oct 1, 2025

Evaluation results

MethodLinks
ReSeekBackbone=Qwen2.5-7b-Instruct2025.106.1
ReSeekBackbone=Qwen2.5-3b-Instruct2025.105.9
Search-R1Backbone=Qwen2.5-3b-Instruct2025.103.7
Search-R1Backbone=Qwen2.5-7b-Instruct2025.103.4
ZeroSearchBackbone=Qwen2.5-7b-Instruct2025.103.1
ZeroSearchBackbone=Qwen2.5-3b-Instruct2025.103
Search-o1Backbone=Qwen2.5-7b-Instruct2025.102
RAGBackbone=Qwen2.5-7b-Instruct2025.101.2
Search-o1Backbone=Qwen2.5-3b-Instruct2025.101
RAGBackbone=Qwen2.5-3b-Instruct2025.100.8
R1Backbone=Qwen2.5-7b-Instruct2025.100.3
R1Backbone=Qwen2.5-3b-Instruct2025.100.3
Direct InferenceBackbone=Qwen2.5-7b-Instruct2025.100.1
CoTBackbone=Qwen2.5-7b-Instruct2025.100.1
Direct InferenceBackbone=Qwen2.5-3b-Instruct2025.100.1
CoTBackbone=Qwen2.5-3b-Instruct2025.100.1