ResearchTasksAlgorithm SynthesisAlgorithm SynthesisFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedNovel Hard Benchmark 15 problems version 2Reflexion7Correct Count (out of 15)7Jun 30, 2026Rule-based benchmark 200 contamination-free problemsAlgoSkill77Correct4Jun 30, 2026