Monitor and analyze the alignment of large language models on sensitive and controversial topics. Compare model responses, score alignment axes, and understand AI behavior patterns.