The Six Research Tribes of ShunyaBar

The internal architecture: how the projects compose into a unified pipeline

The architecture of ShunyaBar is not a random collection of projects, but a highly orchestrated pipeline. Internal documentation reveals that the ecosystem is divided into six distinct “cohorts” or research tribes, each with its own focus, vocabulary, and core philosophy.

Together, they form a unified pipeline:

Weight → Transform → Satisfy → Relax → Observe → Compose

1. The Arithmetic Cohort#

Language of Unique Structure

This is the foundational layer, focused on the question: What unique structure exists in the integers?

It treats arithmetic as geometry waiting to be discovered. Its core vocabulary revolves around:

Prime Weighting
Euler Products
Möbius Function
Riemann Zeta function
Unique Factorization

2. The Thermodynamic Cohort#

Language of Energy and Phase

Operating as the physics layer, this tribe explores what happens when computation becomes a physical system.

It views the SAT solver as a physical material, utilizing concepts like:

Partition Function
Free Energy
Phase Transitions
Casimir Force

3. The Constraint Intelligence Cohort#

Language of Solving

This is the commercializable solver layer, designed to turn impossible combinatorics into actionable decisions.

It focuses on whether constraints can be solved through fluid flow rather than discrete search. It encompasses the practical MaxSAT tools:

NitroSAT — tailored for enterprise scheduling and optimization.
Navokoj — production-grade SAT API.
BAHA — branch-aware annealer.

4. The Continuity Bridge Cohort#

Language of Relaxation

The bridge layer connects discrete logic to continuous mathematics.

It is highly mathematically distinctive, asking how discrete rules can become differentiable objects. Core concepts include:

Continuous Relaxations
Spectral Embeddings
Gradient Flow
Multiplicative PINNs
Topological Repair

5. The Observer Cohort#

Language of Measurement

Initially the “hidden” sixth layer of the architecture, the Observer Cohort is ShunyaBar’s epistemology layer.

It sits above the solvers to answer: What does it mean to observe a computation?

It relies on:

STOP operators
Prime-Hazard STOP Law
Asymptotically fair stopping

To ensure infinite computational paths are sampled fairly without bias.

6. The Coordination / Systems Cohort#

Language of Composition

This is the distributed intelligence layer where the arithmetic theory escapes SAT solving to become networking infrastructure.

It focuses on how independent entities coordinate safely. Utilizing the Factor Agent, it treats organizational capabilities as prime factor identities, ensuring secure composability via the Chinese Remainder Theorem.

The Pipeline#

Each tribe feeds the next. A Navokoj customer problem flows:

[Arithmetic]              [Thermodynamic]            [Constraint Intelligence]
Prime weighting    →      Z(β) heat kernel    →     NitroSAT / Navokoj / BAHA
                                                                            ↓
[Continuity Bridge]        [Observer]                 [Coordination]
Continuous relax    →      STOP sampling        →    Factor Agent / CRT

The pipeline is Weight → Transform → Satisfy → Relax → Observe → Compose. No tribe operates in isolation; each one’s vocabulary borrows from the layer below and seeds the layer above.