ResearchBenchmarksMulti-hop Question Answering on InfoQA Synthetic Multi-hop QA 3-hopFollow98Avg F1 ScoreInfoQA31.4448.726683.28Sep 25, 2025Evaluation ResultsMethodMethodLinksAvg F1 ScoreInfoQAContext Length=0.5k, B...Context Length=0.5k, Backbone=Qwen3-14B2025.0998InfoQAContext Length=Average...Context Length=Average, Backbone=Qwen3-14B2025.0980DirectContext Length=0.5k, B...Context Length=0.5k, Backbone=Qwen3-14B2025.0970DirectContext Length=Average...Context Length=Average, Backbone=Qwen3-14B2025.0934