List your company

Open-domain multi-hop question answering on 2Wiki

75.06Answer F1RAG-RL
47.65654.770561.88568.9995Mar 17, 2025

Evaluation results

MethodLinks
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Min-Max2025.0375.0680.3777.63
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Linear2025.0374.3188.8280.92
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Answerable2025.0372.6179.9876.11
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Max2025.0371.8278.5375.02
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Unanswerable2025.0371.0474.372.63
RAG-RLEval. Setting=Distractor, Curriculum=Min-Max2025.0370.7776.3773.46
Qwen2.5-7B-InstructEval. Setting=Ideal Retrieval, Curriculum=Baseline2025.0370.2954.4961.39
RAG-RLEval. Setting=Distractor, Curriculum=Linear2025.0368.9283.2375.4
RAG-RLEval. Setting=Distractor, Curriculum=Base-Answerable2025.0368.5676.3672.25
RAG-RLEval. Setting=Distractor, Curriculum=Max2025.0367.9977.0872.25
RAG-RLEval. Setting=Distractor, Curriculum=Base-Unanswerable2025.0366.5972.0469.21
Qwen2.5-7B-InstructEval. Setting=Distractor, Curriculum=Baseline2025.0348.7140.1844.03