kazenoko.

I like figuring out how things work.

I’m interested in silicon, neuromorphic computing, software, and research. This is where I collect what I’m making and the questions I’m working through.

Latest from the workbench

01

October 2026–present

Neuromorphic NPU

A first neuromorphic NPU design for Hack Club ASIC. An exploration of computation at the hardware level.

In development↗02

Repository updated · 20 Oct 2026

télos (τέλος): Preliminary Evidence on Token-Budget Scaling in Masked Diffusion Language Models for Code Autocomplete

Autoregressive language models currently dominate code generation, yet they are fundamentally constrained by strict left-to-right generation. Masked Diffusion Language Models (MDLMs) offer a compelling alternative via bidirectional context and non-monotonic iterative decoding.…

Preliminary report↗03

Repository updated · 20 Oct 2026

télos (τέλος): Preliminary Report 2 — Capability Divergence, Positional Encoding, and the Limits of Masked Diffusion Scaling

Following the initial scaling observations in Masked Diffusion Language Models (MDLMs) for code autocomplete, we complete our scaling sweeps across the 12.5M, 25M, and 50M parameter tiers. We identify a fundamental limitation in MDLM capacity allocation: past a critical,…

Preliminary report↗04

Repository created · 19 Aug 2026

iPoint

Documentation from kazenoko-git/iPoint.

Repository documentation↗05

August 2026–present

telos

Research Beyond Autoregressive Models.

Repository documentation↗06

24 Jul 2026

Impressions Apotheosis 2026 — Second Place

Second place in The Architect with Klima at DPS Bangalore North's Impressions Apotheosis.

Awarded↗

Projects ↗Blogs ↗Papers ↗Certifications ↗