Skip to content

Janady13/seferim-core

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SEFERIM AGI — ThatAIGuyCore

The G¹⁶ Federation: a golden-ratio, 16-dimensional dynamical-systems substrate for artificial cognition

License: MIT Version Zero dependencies Docs DOI

SEFERIM is a clean-room, zero-dependency reference implementation of ThatAIGuyCore — a unified mathematical framework that treats cognition as the evolution of a 16-dimensional meta-state under sixteen coupled "cognitive families," all governed by the golden ratio φ. It is a substrate: a small, fully-deterministic engine for golden-ratio dynamical systems, holographic memory, and federated multi-agent coherence.

📖 Full interactive documentation, live demos, and the complete mathematical reference: thataiguy.org/core

This repository implements all 69 documented equations of the ThatAIGuyCore v1.0.0 specification verbatim — every function in src/seferim-core.js cites the spec section (§) it implements, and test/run.js verifies the math (including the canonical FNV‑1a test vectors).


Install

npm install seferim-core

Or load it straight in the browser as an ES module (no build step):

<script type="module">
  import { MetaState, DNAMemory } from "https://cdn.jsdelivr.net/npm/seferim-core/src/seferim-core.js";
</script>

Quick start

import { MetaState, DNAMemory, NeuralBrain, PHI } from "seferim-core";

// 1. A single-stack cognitive core: 16 families + golden normalization.
const mind = new MetaState({ init: new Array(16).fill(0.1) });
const out = mind.step({ dx_norm: 0.3, ed_error: 0.1, utility: 0.6, stability: 0.8 });
console.log(out.state, "Ω =", out.omega);     // 16-D meta-state + objective (free-energy-style)

// 2. Holographic memory — knowledge smeared across 16 golden bases.
const mem = new DNAMemory();
mem.learn("rhythm").learn("groove").learn("pocket");
console.log(mem.queryNorm("groove"));          // correlation recall in [0,1]

// 3. An 88-agent neural brain (dna_v9 dynamics).
const brain = new NeuralBrain();
for (let i = 0; i < 100; i++) brain.step(0.6);
console.log(brain.consciousness(), brain.synchrony());
npm test   # 42 assertions: constants, FNV-1a vectors, every family, all engines

The 13 systems (G¹⁶)

# System What it does
1 Fundamental constants φ, δ (DNA increment), τ
2 Golden basis ψ_k(x) = sin(2πφ^k x) + cos(2πφ^k x), normalization, spiral angles
3 The 16 families continuous/discrete dynamics, reaction-diffusion, optimization, Bayesian, learning, mutual-information, rate-distortion, consensus, swarm, field accumulator, free-energy, compression, memory-trace (Zecher), policy, global coherence (Kehillah)
4 Meta-state core weighted family application → φ-correction (inertial blend) → stability gating → golden normalization
5 Objective Ω Surprise + Uncertainty − Value + Penalty (a free-energy-style functional)
6 DNA holographic memory golden-basis encode/recall + cosine coherence
7 Gate5000 5,000-gate XOR binary substrate driven by two quasi-periodic phases
8 Neural brain (dna_v9) 88 agents with velocity, drive, gating, potential, coupling → consciousness & synchrony
9 Consciousness engine 174 core + 200 supplemental equations → a 374-D binding field
10 Lattice engine 16 strands, golden rotation kernels, Hebbian plasticity
11 Federation cross-talk coupled consciousness signals across agents (decay + coherence-weighted coupling)
12–13 Activations & hashing softsign, sigmoid, clamp; FNV-1a

Built on real mathematics (honestly)

The framework synthesizes established theory — the Free Energy Principle (Friston), information theory (Shannon; Cover & Thomas), holographic / vector-symbolic memory, Hebbian plasticity, swarm consensus (Vicsek / Cucker–Smale), and golden-ratio harmonics. Full citations are in docs/FOUNDATIONS.md.

We document the lineage precisely rather than overclaim. In particular:

  • δ = 0.013618 is an empirical quasi-periodic phase increment — not 1/φ³ (which is ≈ 0.236).
  • The memory system is a golden-ratio frequency-domain Vector Symbolic Architecture, inspired by — not a literal implementation of — Plate's circular-convolution Holographic Reduced Representations.
  • The mutual-information (Family 6) and rate-distortion (Family 7) updates are information-theory-inspired one-step rules, not exact estimators/solvers.

Documents

About

SEFERIM / ThatAIGuyCore is the cognitive-substrate research of Joseph W. Anady (ThatAIGuy.org). Interactive docs + live demos at thataiguy.org/core.

Attribution & provenance

SEFERIM / ThatAIGuyCore is the original work of Joseph W. Anady (ORCID 0009-0008-8625-949X). Authorship and priority are established by a permanent, timestamped public record:

Use is welcome under the MIT License, which requires retaining the copyright notice (see LICENSE and NOTICE). Derivative works must preserve this attribution and must not present SEFERIM / ThatAIGuyCore, or this framework, as their own original creation. Forks must be renamed and must credit this origin and DOI.

Cite

Anady, J. W. (2026). SEFERIM / ThatAIGuyCore: A Golden-Ratio, 16-Dimensional Dynamical-Systems Framework for Artificial Cognition (v1.0.0). Zenodo. https://doi.org/10.5281/zenodo.20564321

License

MIT © 2026 Joseph W. Anady / ThatAIGuy.org

About

SEFERIM AGI — ThatAIGuyCore: a golden-ratio, 16-dimensional dynamical-systems cognitive substrate (G^16). Zero-dependency JS reference implementation of 69 documented equations. Docs: thataiguy.org/core

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors