List your company

Question Answering on HotpotQA 500 QA pairs

38.3Exact Match (EM)Ft. Agent
15.83621.66827.533.332Oct 19, 2025

Evaluation results

MethodLinks
Ft. AgentBackbone=Qwen-2.5-7B-Instruct2025.1038.3
Base AgentBackbone=Qwen-2.5-14B-Instruct2025.1037.5
Ft. AgentBackbone=Qwen-2.5-3B-Instruct2025.1034.7
Naive RAGBackbone=Qwen-2.5-14B-Instruct2025.1032.9
Naive RAGBackbone=Qwen-2.5-7B-Instruct2025.1030.9
Base AgentBackbone=Qwen-2.5-7B-Instruct2025.1028.1
Naive RAGBackbone=Qwen-2.5-3B-Instruct2025.1027.5
Naive RAGBackbone=Mistral-NeMo-8B-Instruct2025.1027.4
Base AgentBackbone=Mistral-NeMo-8B-Instruct2025.1025.7
Base LLMBackbone=Qwen-2.5-14B-Instruct2025.1025.1
Base AgentBackbone=Qwen-2.5-3B-Instruct2025.1022.5
Base LLMBackbone=Mistral-NeMo-8B-Instruct2025.1020.1
Base LLMBackbone=Qwen-2.5-7B-Instruct2025.1019.7
Base LLMBackbone=Qwen-2.5-3B-Instruct2025.1016.7