3 ms·Happy to share this and curious how others are thinking about runtime-level architectures for LLM cognition.by teugent 1y agoHappy to share this and curious how others are thinking about runtime-level architectures for LLM cognition.