Cosmos Research Institute

Where I emulate the multiverse — one cell, one organ, one neuron, one mind at a time.

This repository is the umbrella for a long arc of work: rebuild the universe in software, scale by scale, until what we run on a laptop tells us something true about what we live inside.

The story is split in two halves. Emulation asks what is the world made of? — life, body, brain. Cosmos Intelligence asks what does it mean? — the analytical lens that makes the rest legible.

Each chapter is a self-contained project with its own README, tests, and license.

---

The roadmap

                           ┌────────────────────────────┐
                           │    Cosmos_Intelligence/    │
                           │    UIAS v3.0 — the lens    │
                           └────────────┬───────────────┘
                                        │ interprets
                                        ▼
        ┌───────────────────────────────────────────────────────────┐
        │                       Emulation/                          │
        │                                                           │
        │   Life/  ─►  Human_Body/  ─►  Whole_Brain_Emulation/      │
        │   cells       organism            mind                    │
        └───────────────────────────────────────────────────────────┘

Read top to bottom and the story is: build a world, build a body, build a brain, then build the instrument that understands them.

---

Chapter I — Life   ·   v0.1.0, 75 tests passing

Before there is a body, there is something alive at all.

A multi-layered artificial-life simulation in TypeScript. A 128×128 toroidal grid where every cell carries a 128-bit genome encoding behavioral traits and a small feed-forward neural network. Six phases of complexity — Conway's Game of Life, energy/nutrient ecosystems, evolving genomes with speciation, multi-cell organisms with chemical signaling, neural-network decision-making, and Integrated Information (Φ*) metrics for emergence research. Deterministic from a seed; runs in the browser at ~10 ticks/sec.

cd Emulation/Life
npm install
npm run dev      # interactive simulation in the browser
npm test

→ README  ·  Source

---

Chapter II — Human Body   ·   v0.1.0, 255 tests passing

One thing alive is interesting. One thing alive that breathes, bleeds, digests, defends, and remembers is a different problem.

A multi-scale, integrated human-body simulator in pure Python. Thirteen organ systems (cardio, respiratory, nervous, endocrine, digestive, renal, musculoskeletal, integumentary, immune, …) coupled through a 1 ms cardiac substep and a shared signal bus. Seven pre-built clinical scenarios, REST API, CLI, matplotlib visualization, all 14 phases passing.

from humanbody import HumanBody
body = HumanBody()
for _ in range(60):
    body.step(dt=1.0)
print(body.state_summary())
# → HR=72, MAP=93, PaO2=97, SaO2=0.98, glucose=94, cortisol=15, pH=7.40, ...

→ README  ·  Tutorial  ·  Source  ·  📝 Medium · DEV · GeeksforGeeks

---

Chapter III — Whole Brain Emulation   ·   v0.1.0, 137 tests passing

A body without a brain is a heart looking for a reason.

A connectome-level neural simulation framework. Hodgkin–Huxley channels, AdEx and Izhikevich reduced models, multi-compartment cable equations, STDP and reward-modulated plasticity, glia (astrocytes / microglia / oligodendrocytes), neuromodulation (dopamine, serotonin, NE, ACh), neurovascular coupling, and consciousness metrics (Φ, PCI, GWT) — all validated against biological benchmarks.

from wbe.simulation import SimulationEngine, SimulationConfig
config = SimulationConfig(dt=0.025, duration=1000.0, method="rk4", seed=42)
engine = SimulationEngine(connectome=connectome, config=config)
result = engine.run(duration=1000.0)

→ README  ·  Tutorial  ·  Source

---

Chapter IV — Cosmos Intelligence   ·   UIAS v3.0

Simulating something is not the same as understanding it. This is the part that understands.

Universal Intelligence Analysis System — a 9-layer (L0–L8) analytical pipeline that takes any query (a result from one of the emulators, a research question, a design problem) and runs it through axiomatic foundations, formal logic, theory evaluation, scientific calibration, architectural optimization, pipeline orchestration, intelligence synthesis, trans-systemic awareness, and human-scenario grounding.

• 9 layers (L0–L8) — from axioms to meta-cognition to lived human contexts
• 7 axioms (A1–A7), 5 ontological categories, 8 logic systems, 10 architecture templates
• 6 quality metrics, 4 feedback mechanisms, full process traceability
• CLI (single-shot + REPL) and a dark-themed Flask web UI

cd Cosmos_Intelligence
pip install -e .
uias "Why do complex systems exhibit emergent behavior?"

→ README  ·  Architecture  ·  API

---

What's next

The four chapters above cover the substrate, the body, the brain, and the lens. The arc keeps going outward:

• 🌍 Ecology — multi-organism populations on a shared environment, building on Life/
• 🌌 Stellar dynamics — the same emulation discipline applied at cosmic scale
• 💱 Economic agents — emergent macro behavior from minimal microeconomic rules
• 🧬 Cross-emulator coupling — feed a Whole_Brain_Emulation instance into a Human_Body and let Cosmos_Intelligence watch

The institute earns its name when those last three exist.

---

Getting started

Each chapter is self-contained. Pick one and follow its README.

git clone https://github.com/Oscar-Wu-Po-Wei/Cosmos.git
cd Cosmos

# TypeScript project (Life):
cd Emulation/Life && npm install && npm test

# Python projects (Human_Body, Whole_Brain_Emulation, Cosmos_Intelligence):
cd Emulation/Human_Body && pip install -e ".[dev]" && pytest tests/ -q

CI runs on every push (see .github/workflows/ci.yml) and exercises every subproject.

---

License

Each subproject ships its own LICENSE. Life and Human_Body are MIT.

Contact

Oscar Wu (Po-Wei) — built with the help of Claude Code.