ResearchTasksCross-modal attack detectionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedSkillMutator GPT-5.4 adversarial-oracle (evaluable)GPT-5.486.8Detection Rate9Jun 15, 2026SkillMutator GPT-5.4-mini adversarial-oracle evaluable subset (n=63)GPT-5.488.9Detection Rate9Jun 15, 2026SkillMutator GPT-4o-mini adversarial-oracle evaluable subset (n=48)GPT-5.489.6Detection Rate9Jun 15, 2026
SkillMutator GPT-5.4-mini adversarial-oracle evaluable subset (n=63)GPT-5.488.9Detection Rate9Jun 15, 2026
SkillMutator GPT-4o-mini adversarial-oracle evaluable subset (n=48)GPT-5.489.6Detection Rate9Jun 15, 2026