List your company

Multi-hop Question Answering on HotpotQA (Answer F1, Citation F1, Joint F1)

76.18Answer F1RAG-RL
60.028864.221968.41572.6081Mar 17, 2025

Evaluation results

MethodLinks
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Min-Max2025.0376.1893.1383.81
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Linear2025.0375.6789.3481.94
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Unanswerable2025.0375.2382.9978.92
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Answerable2025.0374.7683.6978.98
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Max2025.0374.2586.2679.81
RAG-RLEval. Setting=Distractor, Curriculum=Min-Max2025.0368.8781.6474.72
RAG-RLEval. Setting=Distractor, Curriculum=Linear2025.0368.7178.5473.3
Qwen2.5-7B-InstructEval. Setting=Ideal Retrieval, Curriculum=Baseline2025.0367.963.2665.5
RAG-RLEval. Setting=Distractor, Curriculum=Base-Answerable2025.0366.1972.8169.34
RAG-RLEval. Setting=Distractor, Curriculum=Max2025.0366.0473.9369.76
RAG-RLEval. Setting=Distractor, Curriculum=Base-Unanswerable2025.0365.2571.1368.06
Qwen2.5-7B-InstructEval. Setting=Distractor, Curriculum=Baseline2025.0360.6536.4745.55