List your company

Multi-hop Question Answering on HotpotQA (RAG and CoT Evaluation)

75.63AccuracyRIRS
42.97451.45259.9368.408Jan 14, 2025

Evaluation results

MethodLinks
RIRSModel=Qwen-Plus2025.0175.6382.8581.8645.37
ChameleonModel=Qwen-Plus2025.0173.1984.178.744.56
RIRSModel=Mixed2025.0169.3380.2773.1753.7
RouterDCModel=Qwen-Plus2025.0167.2472.2675.6445.32
RIRSModel=LLaMA-3.1-8B2025.0158.6977.3364.4239.23
CoTModel=Qwen-Plus2025.0154.58——54.58
ChameleonModel=LLaMA-3.1-8B2025.0150.277.7555.8130.56
RouterDCModel=LLaMA-3.1-8B2025.0147.7474.5150.9238.51
CoTModel=LLaMA-3.1-8B2025.0144.23——44.23