LLM training data mixture optimization breaks when training pools shift — every prior proxy experiment becomes stale.

Chatbots are far more predictable in their responses than you might expect. That's fine for research or coding, but it's a ...