List your company

Open-domain multi-hop question answering on MuSiQue answerable

65.06Answer F1RAG-RL
24.312834.891445.4756.0486Mar 17, 2025

Evaluation results

MethodLinks
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Min-Max2025.0365.0681.5172.37
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Linear2025.0361.173.966.89
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Answerable2025.0357.5367.7562.22
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Max2025.0354.6468.8460.92
RAG-RLEval. Setting=Ideal Retrieval, Curriculum=Base-Unanswerable2025.0354.5368.3760.67
RAG-RLEval. Setting=Distractor, Curriculum=Min-Max2025.0347.1864.4854.49
RAG-RLEval. Setting=Distractor, Curriculum=Linear2025.0344.6859.7951.14
RAG-RLEval. Setting=Distractor, Curriculum=Base-Answerable2025.0341.454.2846.97
Qwen2.5-7B-InstructEval. Setting=Ideal Retrieval, Curriculum=Baseline2025.0341.1658.1648.21
RAG-RLEval. Setting=Distractor, Curriculum=Max2025.0340.9153.0746.2
RAG-RLEval. Setting=Distractor, Curriculum=Base-Unanswerable2025.0338.8452.1544.52
Qwen2.5-7B-InstructEval. Setting=Distractor, Curriculum=Baseline2025.0325.8825.3525.61