Modeling Community Attitude through Reaction Tone: A Human-AI Collaborative Framework for Evaluating LLM Alignment with Linguistic Behaviors in Online Communities

In three linesCARE, an evaluation framework, measures LLMs' ability to simulate authentic community reactions to real-world news. The study reveals a "realism gap": explicit community prompts fail to improve simulation fidelity. Frontier models show divergent behavioral signatures.

Summary generated by Claude — human-verified