ResearchBenchmarksLog-Investigation Query Generation on Audit Simple (test)Follow98.2F1 ScoreHuman97.78497.8929898.108May 21, 2026Evaluation ResultsMethodMethodLinksF1 ScoreDelta ScoreHuman2026.0598.2—Sievecontext_strategy=Drain3context_strategy=Drain32026.0597.8-0.004