Show HN:工業 AI 安全的確定性物理核心(0 項違規對比 59 項)

Show HN:工業 AI 安全的確定性物理核心(0 項違規對比 59 項)

Hacker News·

一篇於 Hacker News 的「Show HN」發布的文章介紹了 AXIOM RE,這是一個專為工業 AI 安全設計的確定性物理核心。該核心聲稱能 100% 遵守親和力定律(Affinity Laws),並達到完美的 500/500 安全率,相較之下其他系統有 59 項違規。

Navigation Menu

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

To see all available qualifiers, see our documentation.

A deterministic physics kernel for industrial AI safety. Enforces 100% compliance with Affinity Laws. (500/500 Safety Rate).

License

Uh oh!

There was an error while loading. Please reload this page.

chachamwise/axiom-re-core

Folders and files

Latest commit

History

Repository files navigation

AXIOM: The Universal Physics Kernel

Image

Image

Image

⚠️ DISCLAIMER: This repository contains no AI, policies, or learning components. > It provides only the deterministic verification kernel (PhysicsKernel) and the constitutional interface (Bridge).
AI agents utilizing this kernel can be found in the experiments/ folder.

AXIOM RE (Reasoning Engine) is a stateless, physics-constrained verification architecture for autonomous industrial systems. It decouples physical laws from control logic, acting as an immutable "Constitutional Layer" for AI agents.

🧠 The Philosophy: "Code is Law"

In critical infrastructure (Water, Energy, Mining), a control error is not a bug—it is a disaster.
Standard AI (RL/NN) is probabilistic and therefore inherently unsafe for zero-tolerance environments.

AXIOM solves this by embedding a Sealed Physics Kernel that verifies every action against algebraic invariants derived from Dimensional Analysis and Affinity Laws. It does not optimize for rewards; it enforces physical feasibility.

🏗️ Architecture

The system operates as a strict three-layer hierarchy. For a detailed engineering breakdown and safety guarantees, see the Formal Scope.

Image

See also: Preprint Abstract & Contributions

🚀 Quick Start

Installation

Usage Example

🧪 Proven Safety

AXIOM demonstrates zero kernel-defined safety invariant violations in the provided experiments.

🛡️ License & Integrity

Licensed under the AXIOM Public License (APL-1.0).

📜 Citation

If you use AXIOM in your research, you must cite the architecture:

Mwise, C. (2026). AXIOM: A Stateless Physics-Constrained Verification Architecture for Safe Autonomous Systems. Aquaflux Tech.

BibTeX:

About

A deterministic physics kernel for industrial AI safety. Enforces 100% compliance with Affinity Laws. (500/500 Safety Rate).

Topics

Resources

License

Uh oh!

There was an error while loading. Please reload this page.

Stars

Watchers

Forks

Releases

  1

Packages

  0

Languages

Footer

Footer navigation

Hacker News

相關文章

  1. Show HN:適用於機器人的 12KB 確定性 AI 核心 (bestbrain-core)

    4 個月前

  2. Show HN:Agent OS – AI 代理的 0% 安全違規保證

    3 個月前

  3. 靜默工作者教學法:無需修改權重即可實現AI對齊

    4 個月前

  4. Show HN:透過代數連續性克服物理AI的5000W運算瓶頸

    4 個月前

  5. Show HN:Remember Me AI (完整發布) – AI 記憶體系統成本大幅降低 40 倍

    4 個月前

其他收藏 · 0