List your company

Multi-hop Question Quality Evaluation on Comparison Questions

98.6Multi-Hop AccuracyHopWeaver
93.71294.98196.2597.519May 21, 2025

Evaluation results

MethodLinks
HopWeaverLLM Engine=Gemini-2.5-flash2025.0598.64.45
2WikiMultiHopQAGeneration Source=Human Datasets (Baselines)2025.0597.64.42
HopWeaverLLM Engine=QwQ-32B2025.0597.44.4
HopWeaverLLM Engine=Qwen3-14B2025.0595.94.36
HotpotQAGeneration Source=Human Datasets (Baselines)2025.0595.64.2
HopWeaverLLM Engine=GLM-4-9B-04142025.0593.94.26