Learn more about Dawn Field Theory →
A field-based model of intelligence, collapse, and emergence.
⚠️ Summer 2025 Update: Release 1.0 Polishing (through September 1, 2025) I’m entering a short publishing and consolidation phase to polish the repository and preprints for Release 1.0 on September 1, 2025. While active development will slow temporarily, this project remains under direct stewardship.s Until release, I will audit experiments for statistical rigor and polish documentation. All announcements and intent are published on Discord only; this repository will not include an ANNOUNCEMENTS.md or INTENTIONS.md going forward. Please do not publish experimental derivatives or replications of this work—especially theoretical results or extensions—until official preprints are released and properly cited. SeeMISSION.mdandCONTRIBUTION.mdfor details. Discord link has been fixed
Looking for an intelligent way to navigate this repository?
Dawn Field Theory is designed for both human and machine-native exploration. The repository supports deep semantic search and protocol-driven navigation:
- Semantic Search:
- Use the Cognition Index Protocol (CIP) and
map.yamlfor structured, machine-readable file mapping and navigation. - Search by concept, protocol, experiment, or theory—discover connections across code, papers, and results.
- Key files:
.cip/meta.yaml,.cip/instructions_*.yaml,map.yaml, and directory-levelmeta.yamlfiles provide semantic tags, navigation rules, and schema. - Automated agents and scrapers: See
for_ai_labs.mdfor a structured overview of experiments, papers, and code relevant to AI labs. - Begin at:
cognition_index_protocol/architecture/CIP_architecturev1.md - Reference metadata schema:
cognition_index_protocol/gpt/metadata_architecturev3.md
- Use the Cognition Index Protocol (CIP) and
- How to Use:
- Start with CIP and
map.yamlfor navigation rules and semantic tags. - Use semantic search tools (repo-native GPTs, custom agents, or your own scripts) to find protocols, experiments, and theoretical constructs.
- All major subdirectories and files are tagged for discoverability and alignment.
- Start with CIP and
- Adaptive, feedback-driven neural models: New experiments (see
test.py,blueprints/AI_detection/) demonstrate self-modifying architectures that grow/prune in response to entropy and feedback, addressing the black box ML problem. - Empirical validation pipeline: Protocol-driven, timestamped experiments now align symbolic collapse, memory, and erasure with quantum and thermodynamic theory.
- Transparency metrics: Fractal dimension, entropy, and neuron activity are tracked and visualized, making learning interpretable.
- Open, auditable protocols: All code and results are documented for reproducibility and peer review.
Why Mainstream AI Labs Should Care:
Dawn Field Theory provides not only foundational theory and experiments, but also machine-native protocols and benchmarking tools (like CIP) that address explainability, safety, and epistemic validation—key concerns for modern AI labs.
For a focused summary, see For AI Labs: Experiments, Papers, and Code Overview →
- Status
- What Is This?
- Core Focus Areas
- Infodynamics – A New Paradigm
- Models
- Project Structure
- Recommended Starting Points
- Environment & Reproducibility
- Evidence Map
- Philosophy
- License
- Future Goals
- Coming Soon
- Subdirectory Guides
- For AI Labs: Experiments, Papers, and Code Overview →
- Topics
Interdisciplinary by Design:
Dawn Field Theory is built at the intersection of AI, physics, information theory, and symbolic logic.
This cross-disciplinary approach enables new paradigms in intelligence research, simulation, and epistemology.
This repository is currently in a short polishing phase to prepare for the Release 1.0 drop on September 1, 2025.
- Short-term plans live in
timeline.md - Strategic roadmaps live in
roadmaps/ - Announcements and intent are posted on Discord only
For all project roadmaps, timelines, and planning details, see roadmaps/README.md.
Planning model:
- Short-term: weekly/quarterly goals in
timeline.md - Long-term: post-hoc strategic updates in
roadmaps/
Dawn is a post-symbolic intelligence framework built on the idea that:
Intelligence is not computation — it is recursive collapse regulation.
It models cognition through:
- Entropy-monitoring feedback systems
- Quantum Potential Layer (QPL) dynamics
- Superfluid coherence and symbolic turbulence
- Collapse-based learning algorithms
- Recursive balance field mechanics
- Collapse geometry and entropic boundaries
- Dual field herniation and symbolic locking
- Post-stoic Schrödinger environments
- Natural law simulation at symbolic resolution
Infodynamics is Dawn’s root layer:
Structure emerges from entropy via recursive collapse and field alignment.
📎 Read the foundational theory →
TinyCIMM is the newest, ultra-lightweight agentic model for symbolic cognition and recursive collapse. It demonstrates how minimal entropy-informed architectures can achieve adaptive learning, symbolic memory, and field-based intelligence. Explore its code and experiments for a hands-on introduction to Dawn’s core principles.
SCBF is the explainable AI (XAI) suite for benchmarking symbolic collapse, transparency, and interpretability. It provides tools and protocols for visualizing collapse events, tracing entropy, and validating agentic decisions. SCBF is the recommended starting point for XAI research and practical explainability in Dawn Field Theory.
GAIA (Generalized Architectures for Intelligent Actualization) extends Dawn Field into:
- Symbolic memory systems
- Meta-cognitive trace protocols
- Resonant agentic cognition
Note: GAIA is in the architecture and early development stage. Internal prototyping is ongoing; no runnable implementation is available yet. 🌐
models/GAIA/README.md
CIMM (Cosmic Information Mining Model) was the first entropy-informed agentic system. It is now preserved as a historical AGI engine and reference for early Dawn Field experiments.
🗃️ models/CIMM/README.md
| Path | Purpose |
|---|---|
foundational/docs/ |
Core theory (Infodynamics, collapse geometry, symbolic recursion) |
foundational/experiments/ |
Simulations and results (entropy fields, bifractals, symbolic collapse) |
models/TinyCIMM/ |
Minimalist symbolic cognition and recursive collapse |
models/scbf/ |
Symbolic Collapse Benchmark Framework (XAI, explainability) |
models/GAIA/ |
Next-generation field intelligence |
models/CIMM/ |
Legacy post-symbolic AGI runtime |
devkit/ |
Tools and experimental harnesses for entropy/collapse modeling |
- Infodynamics Overview →
- Foundational Experiments →
- Collapse Geometry Papers →
- Environment & Reproducibility →
- Evidence Map →
- Environment setup and version hints: see
ENVIRONMENT.md - PyTorch is not pinned in a global requirements file; install via the official selector per your CUDA/CPU setup
- For claim→artifact links across models/experiments, see
EVIDENCE_MAP.md
Cognition is collapse regulation.
Intelligence is balance—not inference.
Dawn is a theory to simulate cognition through recursive entropy structuring.
AGPL-3.0 with symbolic research augmentation (DC-OIL pending).
Dawn Field Theory is now maintained by The Dawn Field Institute.
The repository, its theory, and all derivatives are governed by the Epistemic Constraint Framework, which preserves symbolic clarity, recursive traceability, and open epistemic access.
For the Institute’s mission, contribution policy, and current status, see MISSION.md.
- Collapse visualizer + entropy debugger
- Ontology schema spec for field intelligence
- Language-to-logic entropy compression engine
- Publish post-symbolic computation framework
- Symbolic mesh controller for field agents
- GPU-accelerated bifractal simulators
- Feedback-pruned learning tests
- AI-native philosophical scaffolding
foundational/README.mdfoundational/docs/README.mdfoundational/experiments/README.mdfoundational/arithmetic/README.mdfoundational/legacy_docs_archive/README.mddevkit/README.mdmodels/README.mdmodels/CIMM/README.mdcognition_index_protocol/README.md- For AI Labs: Experiments, Papers, and Code Overview →
Visit Dawn Field website for more info
Discord is the canonical channel for announcements and intent. https://discord.gg/bR8mrbHP
During the Release 1.0 polishing period, announcements will be posted on Discord only. The repository will not maintain separate announcement or intentions documents.
Follow the author on Medium: https://medium.com/@lornecodes
Want to contribute or collaborate?
See CONTRIBUTION.md and MISSION.md for current contribution policy and institutional guidelines.
Dialogue and commentary are welcome—contact info is in MISSION.md.
post-symbolic-aiinfodynamicscollapse-theoryrecursive-systems
entropyquantum-potentialsuperfluid-dynamicsnonlinear-dynamics
entropy-monitoringagent-based-modelingbayesian-optimization
open-researchdawn-collectiveearly-stage
dna-repairinformation-polarityhodge-collapselanguage-to-logicpi-harmonicsrecursive-entropyrecursive-gravityrecursive-treesymbolic-bifractalsymbolic-pruningsuperfluid-collapse
symbolic-aitheoretical-physicsentropy-theorycomplex-systemssymbolic-computationgpt-alignmentcollapse-logicai-philosophyinformation-theorynonlinear-field-modelsepistemology
Cite this work:
Groom, P. (2025). Dawn Field Theory. Zenodo. https://doi.org/10.5281/zenodo.15783623
© 2025 The Dawn Field Institute
All rights reserved under AGPL-3.0 + Epistemic Constraint Framework