3 ms·
Most tokens are word or sub-word level. A token can be "find" or two tokens can be "find" and "ing." To argue that this isn't inherently using the underlying gr
by dvt 3y ago
Most tokens are word or sub-word level. A token can be "find" or two tokens can be "find" and "ing." To argue that this isn't inherently using the underlying grammar rules is plainly wrong. To argue that an LLM trained on English text is capable of "making up words" (in any meaningful sense of that sentence) is even more wrong.
- _ea1k 3y agoYou might think it is wrong, but I think it is a perfectly cromulent case of AI "making up words".
- jstanley 3y agoDoes this satisfy you? Me: give me a random character sequence that is not a pronounceable word ChatGPT: Sure, here's a random character sequence that is not a pronounceable word: "Jkplqxzv"