List your company

Question Answering on Natural Questions (NQ) (EM and F1 Metrics)

45.68Exact Match (EM)RAG-GD
14.760822.787930.81538.8421Apr 9, 2026Apr 16, 2026Apr 24, 2026May 2, 2026May 9, 2026May 17, 2026May 25, 2026

Evaluation results

MethodLinks
RAG-GDBackbone=Llama-3.1-8B, Retriever=E5, K=52026.0545.6855.84
Base adapterBackbone=Llama-3.1-8B, Retriever=E52026.0543.4654.6
RAG-GDBackbone=Qwen-2.5-7B, Retriever=E5, K=52026.0542.9152.71
Base adapterBackbone=Qwen-2.5-7B, Retriever=E52026.0541.7751.22
Vanilla RAGBackbone=Llama-3.1-8B, Retriever=E52026.0540.7252.23
RAG-GDBackbone=Llama-3.1-8B, Retriever=BM25, K=52026.0540.2250.01
Vanilla RAGBackbone=Qwen-2.5-7B, Retriever=E52026.0539.1650.03
Base adapterBackbone=Llama-3.1-8B, Retriever=BM252026.0538.4749.42
SAVERBackbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0437.147.8
SAVERBackbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.0436.845.9
MADBackbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0436.646.9
MADBackbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.0436.244.6
B-2Backbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0435.944.9
B-2Backbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.0435.545.3
Self-RefineBackbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.0434.643.7
RAG-GDBackbone=Qwen-2.5-7B, Retriever=BM25, K=52026.0534.4643.54
B-2Backbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0434.442.9
SAVERBackbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0433.943.8
Self-RefineBackbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0433.543
MADBackbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0433.442.4
CoTBackbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.043341.7
CoTBackbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0432.743.4
Base adapterBackbone=Qwen-2.5-7B, Retriever=BM252026.0532.5741.45
Vanilla RAGBackbone=Llama-3.1-8B, Retriever=BM252026.0531.4140.95
Self-RefineBackbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0431.239.2
Vanilla LMBackbone=LLaMA-3.1-8B, Inference Setting=Zero-shot2026.0429.538.2
CoTBackbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0429.538.6
Vanilla LMBackbone=Qwen-2.5-7B, Inference Setting=Zero-shot2026.0428.236.9
Vanilla RAGBackbone=Qwen-2.5-7B, Retriever=BM252026.0527.6136.66
Vanilla LMBackbone=LLaMA-3.2-3B, Inference Setting=Zero-shot2026.0425.134.8
Query OnlyBackbone=Llama-3.1-8B, Retriever=None2026.0522.4632.51
Query OnlyBackbone=Qwen-2.5-7B, Retriever=None2026.0515.9524.28