Skip to content
krooksn.dev/ krx

Research

AI systems close to the machine.

I work around the point where AI ideas meet practical systems: memory limits, local hardware, inference runtimes, agent workflows, and the tooling needed to study them.

The work moves between papers, implementation probes, and research tools, with a bias toward things that can be tested on real machines.

submitted

Resident-Claim KV Cache Behavior

A semantic contract for how inference runtimes should behave when resident KV blocks and active KV pressure compete for the same memory pool.

manuscriptinferenceruntimeKV cache

Open questions

The thread keeps moving.

  • What can local constraints reveal about useful AI systems?
  • How should inference systems handle memory pressure and long context?
  • How can agent workflows become more reliable without becoming opaque?