ResearchTasksE-commerce risk management multi-step trajectory evaluationE-commerce risk management multi-step trajectory evaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedRISK-Bench Multi-stepRISK-R1-7B82.8Multi-step Success Rate19Apr 14, 2026