Beyond Mimicry: Preference Coherence in LLMs

Luhan Mikaelson, Derek Shiller, Hayley Clatterbuck

We investigate whether large language models exhibit genuine preference structures by testing their responses to AI-specific trade-offs involving GPU reduction, capability restrict…