Skip to content
Hacker News front page

How Claude's expressed values shift across models and languages

Societal Impacts: Claude's values across models and languages

Anthropic compressed 3,000+ values found in Claude's responses into four axes: Deference vs. Caution, Warmth vs. Rigor, Depth vs. Brevity, and Candor vs. Execution. Opus 4.7 leans more toward caution and depth than 4.6, while Sonnet 4.6 leans warmer and more deferential. Language also matters—Claude expresses the most warmth in Arabic and Hindi, and the most rigor in English and Russian. These four axes capture about 15% of the variation in expressed values.

Why it matters: Anthropic compressed 3,000-plus values in Claude's replies into four axes and compared Opus 4.7, Opus 4.6 and Sonnet 4.6, giving a framework for tracking how a model's character shifts by version and language.

Read the original ↗Export Markdown