List your company

Multi-Hop Question Answering on Synthetic Multi-Hop Benchmark 4-hop

1Avg F1InfoQA
0.05360.29930.5450.7907Sep 25, 2025

Evaluation results

MethodLinks
InfoQAContext Length=0.5k, Backbone=Qwen3-8B2025.091
Self Consistency (S-C)Context Length=10k, Backbone=Qwen3-8B2025.090.09