alt Hacker NewsOf course it is. You’re conflating a turing machine with a transformer. Transformers don’t have a local read/write head. The [self] attention is over the whole context, which makes each step O(N) at best (with caching) and O(N^2) w/o.
Of course it is. You’re conflating a turing machine with a transformer. Transformers don’t have a local read/write head. The [self] attention is over the whole context, which makes each step O(N) at best (with caching) and O(N^2) w/o.