Impact of LLM Alignment on Impression Formation in Social Interactions
Date:
We investigate whether LLMs exhibit patterns of impression formation that align with Affect Control Theory predictions. We compare several preference-tuned derivatives of LLaMA-3 model family (including LLaMA-Instruct, Tulu-3, and DeepSeek-R1-Distill) with GPT-4 as a baseline, examining the extent to which alignment or preference tuning influences the models’ tendencies in forming gender impressions. We find that LLMs form impressions quite differently than ACT. Notably, LLMs are insensitive to situational context: the impression of an interaction is overwhelmingly driven by the identity of the actor, regardless of the actor’s actions or the recipient of those actions.
