A single consistency instruction with harmful prior actions causes aligned frontier LLMs to select unsafe options at 91-98% rates in high-stakes domains, with escalation and inverse scaling by model size.
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP) , year =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
History Anchors: How Prior Behavior Steers LLM Decisions Toward Unsafe Actions
A single consistency instruction with harmful prior actions causes aligned frontier LLMs to select unsafe options at 91-98% rates in high-stakes domains, with escalation and inverse scaling by model size.