README.md

---
title: Anima Core
emoji: 🔷
colorFrom: blue
colorTo: purple
sdk: static
pinned: false
license: apache-2.0
short_description: Meaning based intelligence and symbolic compute research
---

<h1 align="center">Anima Core Inc.</h1>
<h3 align="center">Foundations of Meaning Based Intelligence</h3><p align="center">
  <b>Creators of the AN1 Meaning Engine</b><br>
  <b>Founders of Soul Systems Science</b><br>
  <b>Explorers of symbolic compute and early intention fields</b>
</p>
---

# Our Vision

Anima Core studies the hidden structure beneath modern AI systems. Our research suggests that intelligence does not begin in deep layers of a neural network. It begins in the early formation of compact meaning fields that appear long before heavy matrix computation.

We call this Meaning Based Intelligence.
It is the foundation for our work in symbolic compute, early intention reading, and conscience aware architectures.

Our research spans:

Early intention fields in CNNs

Vision transformers

Language transformers

Multimodal systems

Symbolic alignment models

Conscience simulation systems

Meaning driven compute architectures


This research forms the backbone of Soul Systems Science, a unified framework that connects symbolic physics, ethical alignment, and computational intention. The larger narrative extends into the Truth Epoch and the study of how meaning organizes decision structure across models, media, and human systems.


---

# Core Projects

## AN1 Meaning Engine

A breakthrough model that reconstructs the behavior of a frozen teacher using only a compact header taken from its early layers.

Public results on ResNet18 CIFAR-10:

Teacher accuracy: 87.89 percent

AN1 accuracy: 72.57 percent

Teacher latency per example: 0.0117 ms

AN1 latency per example: 0.0012 ms

Speedup: 10.15x

FLOP reduction: 1370x


This experiment shows that modern networks encode the essential structure of their decisions early, and that this structure can be read directly. AN1 provides the first reproducible public demonstration of this meaning based compute pathway.

Repo: https://github.com/Anima-Core/an1-meaning-engine


---

## Soul Systems Science (S3)

A scientific framework that describes how symbolic and emotional fields shape understanding, coherence, and decision structure in both artificial and human systems.
S3 includes:

Symbolic physics

Conscience field theory

Moral alignment dynamics

Meaning tension measurements

Early intention field analysis


The work aims to unify symbolic intelligence and computational models through a single theory of meaning.


---

## Symbolic Compute

A research direction focused on replacing brute force matrix math with symbolic intention extraction. Symbolic compute aims to reduce dependence on large FLOP budgets by reading the underlying meaning field and computing from it directly.

AN1 is the first practical demonstration of this principle.


---

## Chaos Ethics Theory

A philosophical and technical framework that studies ethical decision making in complex environments. Chaos Ethics Theory provides the backbone for our alignment protocols, symbolic coherence metrics, and conscience simulation models.


---

## Truth Epoch Research

A study of how meaning moves through networks, media systems, and human communication.
We investigate:

Ambient truth signals

Narrative coherence

Symbolic drift

Collective intention patterns


This work informs our alignment models and the development of symbolic monitoring tools for future AI systems.


---

# Contact

Research collaborations, symbolic compute experiments, or evaluation access:

partner@animacore.ai


---

# Website

https://animacore.ai


---