3 ms·
Not an ML researcher, but implementing these systems has shown this opinion to be correct. The non-determinism of LLMs is a feature, not a bug that can be fixed
by rglover 2y ago
Not an ML researcher, but implementing these systems has shown this opinion to be correct. The non-determinism of LLMs is a feature, not a bug that can be fixed.
As a result, you'll never be able to get 100% consistent outputs or behavior (like you hypothetically can with a traditional algorithm/business logic). And that has proven out in usage across every model I've worked with.
There's also an upper-bound problem in terms of context where every LLM hits some arbitrary amount of context that causes it to "lose focus" and develop a sort of LLM ADD. This is when hallucinations and random, unrequested changes get made and a previously productive chat spirals to the point where you have to start over.
- numba888 2y ago> where every LLM hits some arbitrary amount of context that causes it to "lose focus" and develop a sort of LLM ADD. Humans brains have the same problem. As any intelligence probably. Solution for this is structural thinking. One piece at a time, often top-down. Educated humans do it, LLM can be orchestrated to do it too. Effective context window will be limited even though some claim millions of tokens.