project and tooling entry points. research interpretation stays in
the notes; implementation detail stays close to each tool.
cpu-first LLM inference and probing engine in Rust. GGUF
loading, quantized CPU paths, hidden-state extraction, and
validation workflows.
rustinferenceggufprobing
local RAG pipeline for querying PDF documents with local
embeddings, vector storage, and inference.
pythonraglocal inference
minimal x86_64 kernel in Rust with paging, heap allocation,
cooperative async execution, ramdisk, and a shell.
rustkernelx86_64no_std