ResearchDatasetsSynthWorldsSynthWorldsFollowBenchmarksTaskDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyTaskDatasetSOTA resultTrendResultsReasoningSynthWorlds 2026 (test)0.48Score15ReasoningSynthWorlds48Reasoning Score (SynthWorlds)15Multi-hop Question AnsweringSynthWorlds-RM26.9Exact Match (EM)10