Comment by positron26
6 months ago
There are no recurrent paths besides tokens. How may I introspect something if it is not an input? I may not.
6 months ago
There are no recurrent paths besides tokens. How may I introspect something if it is not an input? I may not.
The recurrence comes from replaying tokens during autoregression.
It's as if you have a variable in a deterministic programming language, only you have to replay the entire history of the program's computation and input to get the next state of the machine (program counter + memory + registers).
Producing a token for an LLM is analogous to a tick of the clock for a CPU. It's the crank handle that drives the process.
Important attention heads or layers within an LLM can be repeated giving you an "unrolled" recursion.
An unrolled loop in a feed-forward network is all just that. The computation is DAG.
But the function of an unrolled recursion is the same as a recursive function with bounded depth as long as the number of unrolled steps match. The point is whatever function recursion is supposed to provide can plausibly be present in LLMs.
2 replies →
Introspection doesn't have to be recurrent. It can happen during the generation of a single token.