Consider using fastText's word vectors. They have a bunch of languages that come pre sorted by frequency and are sufficient for basic word sense. Perhaps use a

Consider using fastText's word vectors. They have a bunch of languages that come pre sorted by frequency and are sufficient for basic word sense. Perhaps use a LLm to automate some of the disambiguation. https://fasttext.cc/docs/en/crawl-vectors.html https://fasttext.cc/docs/en/crawl-vectors.html https://news.ycombinator.com/item?id=13771292 https://news.ycombinator.com/item?id=13771292 (6 years ago) Aligning the fastText vectors of 78 languages https://github.com/babylonhealth/fastText_multilingual/blob/master/README.md https://github.com/babylonhealth/fastText_multilingual/blob/...