Notes

Paper notes, experiments, and write-ups on KV cache optimization, on-device LLM inference, and NPUs.

• The first notes are on their way.