Hacker News new | ask | show | jobs
by asey 1290 days ago
Models like these don't see words as made of up letters but rather see the whole words (tokens) as a single entity. The result being they're not very good at creating novel (non-memorized) anagrams/palindromes and the like.