ResearchTasksSoftware Engineering Question AnsweringFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSWE-QA ProPredicate-Based Belief State51.4Rubric Judge Accuracy15May 11, 2026SWE-QA ConanLaMR8.71Score6May 18, 2026SWE-QA ReflexLaMR8.15Overall Score6May 18, 2026SWE-QA StreamlinkUnpruned8.74Score6May 18, 2026