ResearchTasksIterative Self-OptimizationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedOPT-BENCH-ML 20 steps 1.0Qwen2.5-3B-Instruct8Win Count19May 12, 2026OPT-BENCH ML environments 10 steps 1.0claude-3-7-sonnet-202502198Win Count19May 12, 2026OPT-BENCH 5 steps 1.0 (ML environments)claude-3-7-sonnet-202502198Win Count19May 12, 2026