Pinned Loading

1. cunningham-chain-search cunningham-chain-search Public

   GPU/CPU Cunningham chain search engine with interactive 2-adic visualizations and analysis tools

   Cuda

2. compartment compartment Public

   Kernel-enforced sandboxing for untrusted processes. Two zero-dependency core tools, one shared profile format, plus an optional BPF-LSM module.

   Shell 2

3. URTB URTB Public

   Encrypted PTY tunnel over ESP-NOW/LoRa radio links or untrusted jump hosts — XChaCha20-Poly1305

   C

4. k-tuplet-search k-tuplet-search Public

   k-tuplet-search

   Cuda

5. int-llm int-llm Public

   Integer-only LLM stack in pure C — run TinyLlama and train a tiny GPT in Q16.48 fixed-point, zero floating point. A reproducible fun experiment.

   C

6. microgpt.c — sub-20ms GPT training a... microgpt.c — sub-20ms GPT training and inference in pure C

   1

   /* Copyright (c) 2026 Nenad Mićić <nenad@micic.be> https://be.linkedin.com/in/nenadmicic

   2

    * SPDX-License-Identifier: Apache-2.0

   3

    *

   4

    * microgpt.c — Minimal GPT training and inference in 655 lines of pure C.

   5

    * Hand-written forward and backward pass, two-phase gradient accumulation.