Downloads · 30 days
0
symbolic-recursion-coherence
dontelightfoot/symbolic-recursion-coherence
symbolic-recursion-coherence is a machine learning model from dontelightfoot. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
<p align="center" <img src="assets/srclogo.jpg" alt="Donte Lightfoot — Symbolic Recursion Coherence" width="280"/ </p
Access
Public
Updated Sep 24, 2026
Repo size
7.6 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.txt3.2 MB · 23%
At a glance
- License
- mit
- Access
- Public
- Created
- Aug 5, 2026
- Updated
- Sep 24, 2026
- SHA
- 2d10cab4
Model card
From the Hugging Face model README
<p align="center">
<img src="assets/src_logo.jpg" alt="Donte Lightfoot — Symbolic Recursion Coherence" width="280"/>
</p>
Symbolic Recursion Coherence (SRC)
Author: Donte Lightfoot
Org: Locked In Safety STL Inc
Support: [email protected]
License: MIT · Status: Open research stack
<p align="center"> <img src="assets/src_banner.jpg" alt="SRC brand banner" width="100%"/> </p>Public mirrors
- GitHub: https://github.com/STLNFTART/symbolic-recursion-coherence
- Hugging Face (this model): https://huggingface.co/dontelightfoot/symbolic-recursion-coherence
- Amyloid package: https://huggingface.co/dontelightfoot/amyloid-beta-42-quantum-prototype
- Engine Space: https://huggingface.co/spaces/dontelightfoot/symbolic-recursive-engine
- Portal Space: https://huggingface.co/spaces/dontelightfoot/symbolic-recursion-coherence
Model Description
Symbolic Recursion Coherence is Donte Lightfoot’s open research stack for measuring and steering coherence in complex systems. It is not a drop-in LLM weight dump (pytorch_model.bin / safetensors chat fine-tune) and not an adapter of Llama, Kimi, DeepSeek, or any other checkpoint. It is a runnable kernel + proof suite + physics DSL + RF telemetry bridge that agents and humans can execute locally, in the Gradio Space, or via MCP.
Do not put pipeline_tag: text-generation, library_name: gradio, or base_model in this file. Those fields are what made the Hub draw a fake model tree (DeepSeek-R1 adapter) and an Inference widget with nothing to serve. Space config (sdk, app_file) belongs on the Space README only.
The stack implements Symbolic Recursion as a Meta-Operator for Coherence in Complex Systems (Primal Logic / Lightfoot kernel):
| Layer | What it does |
|---|---|
| SRC kernel | Recursion, meta-operator, and live sensor coherence. Formulas are not published on this card. |
| DBITS | Dynamic Quantum Bits — 6 executable stability proofs (Jacobian, Lyapunov, gates, PL, C_PL, scale) |
| Intent-encoded algos | Algorithms 1–10 (Primal Logic, JPL, Sovereign Trust, Temporal, Echo, …) modulated by intent vector I |
| Amyloid β-42 | 4-qubit truncated quantum prototype with full circuit embedded |
| PrimalLang v3 | Physics DSL for cymatics, plates, quantum state scripts |
| MCP tools | Agent-facing tools for Cursor / Claude / Grok |
| RF substrate | ESP32 Heltec LoRa TX/RX + RTL-SDR drive — 5 catalog algorithms over the air |
Author profile
Donte Lightfoot builds sovereign control / coherence systems under Locked In Safety STL Inc — spanning symbolic recursion math, PrimalLang physics scripts, DBITS proofs, Amyloid quantum exhibits, and live RF telemetry (ESP32 ↔ RTL-SDR ↔ Databricks). This Hub repo is the public research face of that stack.
What ships in this repo
- Live sensors (Open-Meteo + NOAA K-index) rolled into one coherence reading
- Kernel integrators with coherence diagnostics
- Chat routed through HF Inference Providers (Llama 3.1 8B primary) with kernel tool context
- MCP server for autonomous agent tool use
- PrimalLang v3 cymatics / quantum / plate-physics runners
- Amyloid β-42 package (Hamiltonian, statevector, samples, full circuit)
- DBITS proofs + intent-encoded algorithm suite
Amyloid beta-42 (open exhibit)
from symbolic_recursion.models.amyloid_beta import AmyloidBeta42Model
m = AmyloidBeta42Model()
print(m.run()["status"]) # package verification; formulas are not printed
Claim boundary: 4-site truncated Aβ42-labeled quantum prototype. Not a full biological model. Not medical advice.
Package: models/amyloid_beta/ · Portfolio: models/amyloid_beta/SCHOLARSHIP_PORTFOLIO.md
Numeric constants and operator formulas are not printed on this card.
Intended Uses
- Coherence diagnostics — recursion integrity, collapse risk, sensor coupling
- Live environment — temperature, pressure, geomagnetic, and wind, combined in the kernel
- Physics simulation — PrimalLang chladni plates, eigenfreq, modal response, quantum gates
- Agent tooling — MCP server for autonomous agent-to-agent compute deals
- Research — 20-operator benchmark table (gradient, Fourier, wavelet, etc.)
Out-of-Scope Uses
- Not a medical, financial, or safety-critical decision system
- Sensor values are live public feeds — not calibrated instrumentation
- Chat responses depend on upstream LLM providers; kernel tools ground but do not guarantee factual accuracy
How to Use
Space UI
Interactive engine: https://huggingface.co/spaces/dontelightfoot/symbolic-recursive-engine
Static portal (docs and links): https://huggingface.co/spaces/dontelightfoot/symbolic-recursion-coherence
Open the engine Chat tab and ask about:
- Live sensor coherence
- Coherence retention and recursion integrity
- Cymatics at 528 Hz (or any frequency)
- 3IA Atlas fusion
Use Kernel Tools and PrimalLang tabs for direct tool execution. Chat falls back to the local kernel when HF_TOKEN is not set as a Space secret.
MCP (agents)
python -m symbolic_recursion.mcp.server
| Tool | Purpose |
|---|---|
live_sensors | Current Q(t) + instant Dx |
coherence_analysis | Retention score + diagnosis |
primallang_cymatics | Chladni plate physics at Hz |
primallang_fuse_q | Fuse sensor series into PrimalLang |
atlas_fusion | 3IA recursive memory kernel |
chat_with_kernel | Full inference with auto tool selection |
Local API
pip install -e .
uvicorn symbolic_recursion.server:app --port 8080
| Endpoint | Description |
|---|---|
GET /health | Kernel + engine status |
POST /api/chat | Tool-augmented inference |
POST /api/primallang/run | Run .primal scripts |
POST /api/primallang/cymatics | Cymatics demo |
GET /api/tools/{name} | Direct kernel tool access |
Chat routers (not base models)
Optional Hugging Face Inference Providers. They are not base_model parents of this repo:
| Model | Role |
|---|---|
meta-llama/Llama-3.1-8B-Instruct | Primary |
moonshotai/Kimi-K2-Instruct-0905 | Fallback |
deepseek-ai/DeepSeek-R1 | Fallback |
openai/gpt-oss-120b | Fallback |
Embeddings: sentence-transformers/all-MiniLM-L6-v2 (semantic coherence scoring)
Data Sources
| Source | URL | API key |
|---|---|---|
| Open-Meteo | api.open-meteo.com | None |
| NOAA K-index | services.swpc.noaa.gov | None |
No Hugging Face datasets are bundled. Q(t) is computed live at inference time.
Evaluation Metrics
| Metric | Description |
|---|---|
coherence_retention | Correlation-based retention of Q(t) through Dx integration (0–1) |
semantic_coherence | Embedding distance drift across chat session turns (0–1) |
20-operator benchmark table available via benchmark_operators MCP tool or GET /api/benchmarks.
Environment Variables
| Variable | Purpose |
|---|---|
HF_TOKEN | Hugging Face Inference API (required for LLM chat on Space) |
HF_CHAT_MODEL | Override primary model (default: Llama 3.1 8B) |
HF_EMBED_MODEL | Embedding model for semantic coherence |
GATEWAY_URL | Optional physics gateway (http://localhost:3000) |
SENSOR_LAT / SENSOR_LON | Default sensor coordinates |
Citation
@software{symbolic_recursive_engine,
title = {Symbolic Recursive Engine},
author = {Donte Lightfoot},
year = {2026},
url = {https://huggingface.co/dontelightfoot/symbolic-recursion-coherence},
note = {Interactive engine: https://huggingface.co/spaces/dontelightfoot/symbolic-recursive-engine}
}
License
MIT — see repository for full terms.
- License
- mit
- Languages
- en
- Created
- Aug 5, 2026
- Updated
- Sep 24, 2026