upvote

points

by mfro13 hours ago |
comments
upvoteby steve-atx-760013 hours ago|
[-]
Inference from an LLM is O(tokens^2)
reply
upvoteby halJordan10 hours ago|
parent|
[-]
Only in the naive implementations of attention
reply