Ask HN: GPT non-deterministic, feature or bug?

Seems like GPT non-determinism is mostly from sparse MoE. Imagine if you use a DeepMind approach of tree traversal for the most promising paths, it can generate deterministic LLM responses.