List your company

Question Answering on TriviaQA 500 QA pairs

55.4Exact Match (EM)Naive RAG
30.4436.9243.449.88Oct 19, 2025

Evaluation results

MethodLinks
Naive RAGBackbone=Qwen-2.5-14B-Instruct2025.1055.4
Ft. AgentBackbone=Qwen-2.5-7B-Instruct2025.1054.9
Base AgentBackbone=Qwen-2.5-14B-Instruct2025.1054.1
Base LLMBackbone=Qwen-2.5-14B-Instruct2025.1050.9
Ft. AgentBackbone=Qwen-2.5-3B-Instruct2025.1049.5
Naive RAGBackbone=Mistral-NeMo-8B-Instruct2025.1049.5
Naive RAGBackbone=Qwen-2.5-7B-Instruct2025.1048.8
Naive RAGBackbone=Qwen-2.5-3B-Instruct2025.1045.9
Base AgentBackbone=Qwen-2.5-7B-Instruct2025.1045.5
Base AgentBackbone=Mistral-NeMo-8B-Instruct2025.1043.1
Base LLMBackbone=Mistral-NeMo-8B-Instruct2025.1041.5
Base LLMBackbone=Qwen-2.5-7B-Instruct2025.1039.2
Base AgentBackbone=Qwen-2.5-3B-Instruct2025.1035.7
Base LLMBackbone=Qwen-2.5-3B-Instruct2025.1031.4