ResearchTasksReal-world Query EvaluationReal-world Query EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedWildBench+RL (Skywork-Reward-V2-Llama-3.1-8B)71.5WildBench Accuracy14Mar 4, 2026