ResearchTasksPuzzle ReasoningPuzzle ReasoningFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedReasoning Gym (test)AIPO17.8Avg@420May 12, 2026FrozenLakeo4-mini82Success Rate17May 6, 2026