CITATION.cff

cff-version: 1.2.0
message: "If you use this software in research, please cite it as below."
title: "keeloq-python: Algebraic + neural differential cryptanalysis of reduced-round KeeLoq"
abstract: >-
  A 2026 Python 3 modernization of a 2015 SAT-based cryptanalysis project against the
  KeeLoq block cipher. Ships two pipelines: an XOR-aware ANF encoder + CryptoMiniSat
  algebraic attack that recovers a 64-bit key at 64 rounds with 4 plaintext pairs and
  zero hints in under a second on an RTX 5090 (vs. 14 hours for the 2015 baseline at
  160 rounds / 25 hints / 2 pairs), and a Gohr-style PyTorch ResNet-1D-CNN neural
  differential distinguisher hybridized with the SAT pipeline via Bayesian
  one-bit-per-round key recovery. Includes a documented negative-result analysis of
  the distinguisher's architectural signal horizon on deeper-round KeeLoq.
authors:
  - family-names: Humphreys
    given-names: Cahlen
    alias: cahlen
type: software
license: MIT
repository-code: "https://github.com/cahlen/keeloq-python"
url: "https://github.com/cahlen/keeloq-python"
keywords:
  - cryptanalysis
  - keeloq
  - sat-solver
  - cryptominisat
  - differential-cryptanalysis
  - neural-cryptanalysis
  - gohr
  - pytorch
  - algebraic-cryptanalysis
year: 2026