← Back to context

Comment by duskdozer

17 hours ago

The difference is that LLM output is very nondeterministic.

And because of that, we check in the generated code, not the high-level abstraction. So to understand your program, you have to read the output, not the input.

It depends. Temperature is a variable. If you really need determinism, you could build a LLM for that. Non-determinism can be a good feature though.

  • How would you do that? If it's possible, it seems strange that someone hasn't done it already.