ResearchBenchmarksLog-Investigation Query Generation on SSH Simple (test)Follow99.4F1 ScoreHuman94.82496.01297.298.388May 21, 2026Evaluation ResultsMethodMethodLinksF1 ScorePerformance DeltaHuman2026.0599.4—Sievecontext_strategy=Drain3context_strategy=Drain32026.0595-0.044