Study finds frontier models drift toward Japan after fine-tuning on cultural prompts

ImaginaryRea1ity · reddit · 2026-07-29

A Cardiff University and HiTZ study tested 31,680 culture-related prompts across 24 languages on frontier models including GPT, Gemini, and Claude.

The researchers found that 6 of 8 models drifted toward Japan when asked about dances, festivals, and daily rituals.
The bias appears to emerge after supervised fine-tuning, not in raw pretraining data.
Base models distributed references more evenly, while fine-tuned models leaned heavily toward Japan and the US.
The authors argue Japan becomes a “safe” default: recognizable, globally familiar, and less likely to trigger controversy.

More from AGI Musings