Latin Extended scores highest because phonetic extensions are deliberately designed to resemble their Latin base forms. Mathematical Alphanumeric Symbols dominate the dataset (806 of 1,418 pairs) but score low because ornate mathematical letterforms (script, fraktur, double-struck) look nothing like plain Latin in a different font. Arabic scores lowest: the letterforms are structurally different from Latin even when confusables.txt maps them as confusable.
▲图片来源：https://x.com/Prathkum/status/2027069198091071664/photo/2，更多细节参见91视频
pixels checkpoint restore。搜狗输入法2026对此有专业解读
Scientists say the findings are a warning about the kinds of chemical that people are exposed to in the environment.