Reasoning-Token Spikes Under Prompted Untruthful Responding in Large Language Models
[Sztuczna inteligencja] Autorzy: Maverick Morales, Tomáš Dominik, Vermut Gao i 4 więcej
Monitoring the chain-of-thought of reasoning artificial intelligence (AI) models remains a key approach to detecting deception and other forms of misbehavior in such models. However, semantic chain-of-thought monitoring depends on reasoning traces being legible and sufficiently faithful to the underlying computations that produced the model's behavior, not to mention accessible. Moreover, there is increasing evidence that chain-of-thought outputs may soon become illegible or unfaithful, if they even remain accessible. Based on cognitive load theory, we investigate a lower-bandwidth signal -- the number of reasoning tokens generated -- which does not require access to the content of the reasoning trace. Three reasoning-capable large language models answered 210 multiple-choice questions -- across analytic, descriptive, and normative reasoning types as well as moral and non-moral domains …
- Kategoria arxiv
- cs.AI
- Kategoria
- Sztuczna inteligencja