Symmetry Defeats AuditingFollowAboutWe demonstrate an attack on Introspection Adapters (Shenoy et al., 2026).Nick Merrill, Zeke Medley · 2026Related benchmarksTaskDatasetTaskDatasetResultRankResults ↓Malicious behavior detectionMalicious fine-tunes Llama-3-70B-Instruct (test)IA (pre)1007