ResearchTasksMulti-hop tool-use evaluationFor AI assistantsMarkdownContent indexMulti-hop tool-use evaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedMulti-hop 450 tasks (test)PROBE&PREFILL89Accuracy18May 12, 2026