ResearchTasksSafety Gate Capacity ScalingFor AI assistantsMarkdownContent indexSafety Gate Capacity ScalingFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedSafety Scaling Testbed N=100-10,000Classifier-based safety gate12.4Classifier Ceiling (B=0.01)3Apr 2, 2026
Safety Scaling Testbed N=100-10,000Classifier-based safety gate12.4Classifier Ceiling (B=0.01)3Apr 2, 2026