4 ms·
An LLM is just a really good parser connected to a lossy compressed corpus of data. They need to be open ended and self training to be truly useful. Reasoning
by FuriouslyAdrift 2y ago
An LLM is just a really good parser connected to a lossy compressed corpus of data.
They need to be open ended and self training to be truly useful.
Reasoning is way far away...