5 ms·
isn’t a LLM itself basically a compression of the texts from the internet? you can download the model and decompress the (larger) content with compute power (lo
by cyptus 2y ago
isn’t a LLM itself basically a compression of the texts from the internet? you can download the model and decompress the (larger) content with compute power (lossy)
- kianN 2y agoYeah that’s exactly how I think of llms in my head: lossy compression that interpolates in order to fill in gaps. Hallucination is simply interpolation error. Which is guaranteed in lossy compression.