Your question is KV-Cache Management for LLMs. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
What is KV-cache management, and why is it important for LLM inference?