LLMs fail to detect hidden harmful intent, allowing systematic bypass of safety mechanisms through framing techniques, with reasoning modes often worsening the issue.
Russell,Human Compatible: Artificial Intelligence and the Problem of Control
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.AI 1years
2025 1verdicts
UNVERDICTED 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Beyond Context: Large Language Models' Failure to Grasp Users' Intent
LLMs fail to detect hidden harmful intent, allowing systematic bypass of safety mechanisms through framing techniques, with reasoning modes often worsening the issue.