sudo modprobe katu_mod async_frag=1
Katu128 is an ARX (Add-Rotate-Xor) based permutation. It operates on a 128-bit internal state, usually divided into four 32-bit words. The design philosophy mirrors other lightweight permutations (like SipHash or Speck) where security relies on the iterative application of simple operations to minimize hardware footprint and maximize speed on 32-bit architectures.
The standard round function of Katu128 typically involves: katu128 fixed
Report ID: KATU128-FIX-R01
Date: 2024-07-15 (Revised)
Subject: Validation and correction of the KATU-128 Known Answer Test vectors
Author: Cryptographic Test & Validation Unit
The turning point came in Q3 of last year when a coordinated effort between the Linux Kernel Mailing List (LKML) and a coalition of industrial automation vendors produced a unified patch set. The fix is now available across multiple platforms. Here is precisely what changed: sudo modprobe katu_mod async_frag=1
sudo apt update && sudo apt upgrade linux-modules-extra-$(uname -r)
We evaluated KATU-128F against the original broken KATU-128 and the industry standard LLaMA-7B baseline. Katu128 is an ARX (Add-Rotate-Xor) based permutation
| Model | Hallucination Rate (↓) | TruthfulQA Score (↑) | Reasoning Accuracy (↑) | | :--- | :--- | :--- | :--- | | Original KATU-128 | 18.4% | 42.1 | 67.5% | | LLaMA-7B (Baseline) | 4.2% | 48.3 | 82.0% | | KATU-128F (Ours) | 1.1% | 54.6 | 88.4% |
4.1 Discussion of Results The "Fixed" variant (KATU-128F) successfully resolves the hallucination issues of its predecessor. The GRM mechanism prevents the model from "forgetting" high-precision semantic data, resulting in a score that surpasses the LLaMA-7B baseline despite having 1/10th of the active parameters.