ResearchTasksInstruction Following EvaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSelf-instruct EvalBPO56.7Win Rate (A)5Feb 26, 2026Vicuna EvalBPO63.8Win Rate (A)5Feb 26, 2026IFEval (random subset of 50 prompts)DIRECTER85.9Task Fidelity3Mar 10, 2026IFEval (dev)GPT-492Accuracy3Feb 26, 2026Page 2 of 2PreviousNext