ResearchBenchmarksHallucination Detection on RAGTruth Span-level baseless vs cleanFollow64.1AUCGASP61.91662.48363.0563.617Jul 5, 2026Evaluation ResultsMethodMethodLinksAUCGASPScorer=SmolLM2-1.7B, P...Scorer=SmolLM2-1.7B, Protocol=leakage-clean2026.0764.1GASPScorer=Qwen2.5-1.5B, P...Scorer=Qwen2.5-1.5B, Protocol=leakage-clean2026.0763.8GASPScorer=Qwen2.5-0.5B, P...Scorer=Qwen2.5-0.5B, Protocol=leakage-clean2026.0762