List your company

Single-hop Question Answering on NQ, PopQA, and AmbigQA (test)

59.45F1 (NQ)JADE
15.853227.171638.4949.8084Jan 29, 2026

Evaluation results

MethodLinks
JADEBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Joint Agentic Dynamic Execution2026.0159.4550.268.9459.53
MMOA-RAGBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=RL-Based (Static Modular Workflow)2026.0155.4450.2168.0257.89
RRRBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=RL-Based (Static Modular Workflow)2026.0154.650.4665.4156.82
BGMBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=RL-Based (Static Modular Workflow)2026.0154.2149.5165.9756.56
Search-R1Backbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Agentic Search (Adaptive Workflow)2026.0152.5746.9865.2554.93
Vanilla RAGBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Standard Baseline, RAG Configuration=Simple RAG2026.0140.642.7456.246.51
MAO-ARAGBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Agentic Search (Adaptive Workflow)2026.0136.8241.8547.0341.9
Adaptive RAGBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Agentic Search (Adaptive Workflow)2026.0136.5235.5945.3239.14
LLM w/o RAGBackbone=Qwen2.5-7B-Instruct, Workflow Paradigm=Standard Baseline, RAG Configuration=None2026.0117.5314.9323.5318.66