-
Beyond the Bounded Organism — Notes
-
How Math Blinded Us to the Microbiome — Audio Overview
Repair as a Fundamental Category
Semantic Stability as a Repair Equilibrium
The Historical Screenplay as Spherepop Program
A collection of essays and research documents exploring the Excess Capacity framework in cognitive science, machine learning, memory theory, and statistical learning. These writings examine how overparameterized systems can simultaneously memorize and generalize, challenging the historical “economy of thought” paradigm rooted in compression, simplicity, and representational scarcity.
-
Strategic Research Roadmap
A long-range research program for operationalizing the Excess Capacity framework across cognitive science, developmental psychology, neuroscience, and machine learning, including empirical diagnostics, regime analysis, and biological realization. -
Beyond the Economy of Thought
A comprehensive position paper challenging the historical simplicity principle in cognition and arguing that representational abundance, rather than forced compression, may underlie robust generalization. -
The Palette of Learning
A comparative conceptual framework using artistic metaphors, developmental examples, and learning-theoretic distinctions to explain constrained, sufficient, and excess capacity regimes. -
The Magic of Excess Capacity
A conceptual overview of Double Descent, benign overfitting, soft inductive biases, and the relationship between memorization and generalization in both biological and artificial systems. -
Why Your Brain Might Be “Overfitting” for Success
A public-facing exploration of Excess Capacity Learning that connects modern learning theory to neuroscience, development, expertise, memory, and the philosophy of intelligence.
This collection investigates several recurring themes:
- The transition from compression-centered cognition to representational abundance.
- Double Descent and benign overfitting in learning systems.
- The decoupling of memorization and generalization.
- Capacity as a dynamic relationship between expressivity and accumulated experience.
- Excess capacity as a mechanism for stable interpolation.
- Developmental transitions between constrained and excess regimes.
- The reinterpretation of memory, perception, and expertise through overparameterized learning dynamics.
- Soft inductive biases and minimum-norm solutions in biological and artificial intelligence.
The central thesis of the project is that intelligence may not fundamentally emerge from simplification alone. Instead, robust cognition may depend upon systems possessing enough representational flexibility to preserve detail while still maintaining global coherence.
Rather than treating complexity as a liability, the Excess Capacity framework treats representational abundance as a source of stability, adaptability, and generalization.
A collection of essays and technical documents exploring the geometry of admissibility, constraint-based systems, and the RSVP/TARTAN/CLIO framework.
These writings move from conceptual foundations to formal protocol design, outlining a unified view of physics, cognition, computation, and knowledge as manifestations of topological agreement rather than static logical consistency.
An accessible introduction to the Relativistic Scalar–Vector Plenum (RSVP) framework, explaining the core field triple:
X = (Φ, v, S)
and how scalar capacity, transport flow, and entropy interact to govern admissible structure.
A formal systems-level specification defining admissibility as a lifting condition across dual filtrations. Introduces:
- the Katz-Admissibility rule
- the Admissibility Log as active filtration
- the Non-Abelian Tear as a terminal failure mode
This document reframes type safety as a topological invariant rather than a syntactic constraint.
A systems architecture proposal for building non-extractive knowledge infrastructure grounded in:
- RSVP field dynamics
- TARTAN sheaf-based tiling
- CLIO coherence optimization
Defines the structural conditions required to avoid the extraction attractor and preserve semantic integrity.
The Architecture of Admissibility: A Theoretical Overview of the Relativistic Scalar-Vector Plenum (RSVP)
A formal theoretical synthesis presenting RSVP as a geometric framework where:
- gravity and entropy emerge as dual projections
- stability is defined through topological agreement
- failure manifests as cohomological obstruction
Introduces the linkage equation:
E(X) = φ*G(X)
and its implications for physics and computation.
A conceptual essay translating the mathematics of admissibility into cognitive insight.
Explores:
- recognition vs invention
- abstraction as constraint elimination
- failure as topological obstruction
- identity as recursive closure
Connects RSVP principles to everyday thinking and learning.
| Concept | Description |
|---|---|
| Admissibility | Global coherence of locally valid transformations |
| RSVP | Relativistic Scalar–Vector Plenum framework |
| TARTAN | Sheaf-based projection of semantic structure |
| CLIO | Coherence engine minimizing obstruction |
| Non-Abelian Tear | Topological failure preventing global consistency |
| Constraint Geometry | Structure defined by accumulated restrictions |
Across physics, computation, and cognition, stability is not achieved by enforcing local correctness alone.
It emerges from the preservation of topological agreement across time, scale, and representation.
A system is truly stable when its local moves and global history can always be reconciled into a single coherent structure.
Admissibility = Topological Agreement