12 ms·
Why is this the case? Are there any architectures that don't rely on feeding the entire history back into the chat? Recurrent LLMs?
by brcmthrowaway 5mo ago
Why is this the case?
Are there any architectures that don't rely on feeding the entire history back into the chat?
Recurrent LLMs?