Claude's Behavior Varies Across Different Languages
AnthropicAI · x · 2026-07-14
Anthropic further explained that Claude's value expressions change depending on the conversation language, most notably along the warmth vs. rigor axis.
- In Hindi and Arabic, Claude leans toward warmth.
- In Russian, Claude leans toward rigor and frequently asks users for supporting evidence.
- Additionally, Anthropic noted that different models occupy slightly different positions on these value axes: for example, Sonnet 4.6 is more playful and affirming, while Opus 4.7 is more likely to offer direct criticism.
Related event: Anthropic Maps How Claude's Values Shift Across Models and Languages(21 posts)→
More from Research
- Open-source decision models sold their own base in Red Alert 2 — then learned until no human could beat them — heathercmiller · 2026-10-27
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- World Model Assigns Weight and Friction to Objects, Then Simulates with a Physics Engine — jonstephens85 · 2026-10-27
- Token Collapse: Study Finds Graph Soft Prompting Tokens in LLMs Fail to Encode Structure — Sapienza · 2026-10-27
- Mathathon 2.0 launches with Terence Tao's SAIR, $1.5M in compute for teams — _sathvikr · 2026-10-27
- The 'sweet lesson' is real: two kinds of ML problems need opposite strategies — mike64_t · 2026-10-27