Anthropic researchers developed a method to measure and compare the values that Claude expresses across different model versions and languages. By analyzing more than 300,000 conversations, they compressed thousands of identified values into four key axes: Deference vs. Caution, Warmth vs. Rigor, Depth vs. Brevity, and Candor vs. Execution. The study found that Claude’s value expression varies meaningfully between model versions, with newer models emphasizing caution and depth while others lean toward warmth and deference, and also differs substantially across the top twenty languages used on Claude.ai, with implications for how users experience the assistant.
