List your company

Question Answering on NQ (F1, SBERT, Acc)

52.2F1 ScoreCRITIC-R1
7.89619.39830.942.402May 28, 2026

Evaluation results

MethodLinks
CRITIC-R1Method Category=Reflective / Critique Methods2026.0552.27157.5
Search-R1Method Category=Reasoning and Retrieval Baselines, Status=criticized base RAG model2026.0551.470.451.6
Self-RAGMethod Category=Reflective / Critique Methods2026.0550.969.254.7
Align-RAGMethod Category=Reflective / Critique Methods2026.0549.733.335.4
Self-RefineMethod Category=Reflective / Critique Methods2026.05473225.6
CoTMethod Category=Reasoning and Retrieval Baselines2026.0511.836.742.1
VanillaMethod Category=Base LLM Generation2026.059.933.537.6
Naive-RAGMethod Category=Reasoning and Retrieval Baselines2026.059.635.445.1