Skip to main content
Boundary First Labs
GLOBAL OBJECT NAVIGATIONSearch the Lab
Quick navigation61 objects · 26 declared relationships · 25 pages
↑↓ MoveEnter OpenEsc Close

PUBLICATIONS

Read the work. See what it claims and what still needs to be tested.

Boundary First Labs publishes working papers, research notes, technical reports, formal specifications, experiment reports, reference implementations, and other research artifacts.

Each public record should make its status, claim limits, evidence, open questions, and correction path easier to inspect. A polished document is not presented as stronger evidence than the work behind it.

PUBLICATION PRINCIPLEA reader should be able to tell what is established, what is still being tested, what evidence matters, and what could change the conclusion.

PUBLICATION INDEX

Selected research artifacts, with status and limits visible.

This is a curated first look at current publication records. Being shown here does not mean a paper has been published, peer reviewed, proven correct, or approved for release.
CATALOGCURATED PUBLIC SELECTION
SELECTED RECORDS8 BOUND
LAB SOURCE REVISIONb54314584545
REG-PUBLICATION-SEQUENCEPublication Sequence

Governance / dependency state

CURRENT SOURCE SCALE
45 records
WHAT THIS SOURCE CONTROLS
publication planning, sequence, and dependency-control state only; no scientific validity, readiness guarantee, publication promotion, truth, novelty, or release authorization
Inspect source record
REG-PUBLICATION-GRAPHGlobal Publication Graph

Routing / readiness projection

CURRENT SOURCE SCALE
149 records
WHAT THIS SOURCE CONTROLS
operational publication routing, dependency, and readiness analysis only; no local scientific authority, publication promotion, institutional promotion, or source-state replacement
Inspect source record
REG-PUBLICATION-SOURCESPublication Source Registry

Identity / provenance routing

CURRENT SOURCE SCALE
68 records
WHAT THIS SOURCE CONTROLS
source identity, provenance, import decision, local/canonical correspondence, and routing metadata only; local publication-control artifacts remain authoritative and registration does not promote publication state
Inspect source record
PUBLICATION STATUS BOUNDARYThis page is a curated public view of current publication records. Being listed, sequenced, or marked ready does not mean a paper has been published, peer reviewed, proven correct, or authorized for release.

MORE FROM THE PUBLICATION FIELD

Follow the machinery across established domains.

Computer science leads this supporting layer, followed by mathematics and physics calibrations. Selection is editorial; stage and readiness remain source-governed.
Computer science01

Database theory

Relational Normalization and Lossless Join as Admissible Forgetting

Decomposition forgets structure only when dependencies permit exact reconstruction without spurious tuples. Lossless join becomes a precise reconstruction obligation rather than a loose analogy.

STAGE AREADINESS 4
NormalizationLossless joinSource-read priority
PUBLICATION_GRAPH.json#cs_database_normalization
Computer science02

Automata / formal languages

Myhill–Nerode as Admissible Forgetting

State distinctions may be forgotten exactly when no admissible future observation distinguishes them. Minimal DFA construction becomes an exact observational quotient preserving language behavior.

STAGE AREADINESS 3
AutomataState quotientFuture observation
PUBLICATION_GRAPH.json#cs_myhill_nerode
Computer science03

Algorithms / optimization

Dynamic Programming as Sufficient State Retention

History may be forgotten only when the retained state preserves every distinction relevant to future optimal continuation. DP state design is therefore a concrete sufficiency problem.

STAGE AREADINESS 3
StateSufficiencyBellman principle
PUBLICATION_GRAPH.json#cs_dynamic_programming
Computer science04

Compilers / programming languages

Compiler Pipelines as Representation-Preserving Transport

Each intermediate representation is a representation boundary; compiler passes transport state while preserving declared semantics and forgetting only distinctions irrelevant to the target obligation.

STAGE AREADINESS 3
IRTransportSemantics preservation
PUBLICATION_GRAPH.json#cs_compilers
Computer science05

Distributed data processing

MapReduce as Boundary Formation, Equivalence, and Closure

Shuffle forms intermediate equivalence classes, reduce closes each class, and combiner legality depends on composable local closure. Key design determines what the computation preserves or forgets.

STAGE AREADINESS 3
Distributed systemsEquivalence classesClosure
PUBLICATION_GRAPH.json#cs_mapreduce
Computer science06

Graph algorithms

Dijkstra’s Algorithm as a Moving Closure Frontier

A settled vertex crosses a frontier when no remaining admissible path can improve its distance. Nonnegative weights are the condition that makes this finalization monotone and lawful.

STAGE AREADINESS 3
Graph algorithmsInvariantClosure frontier
PUBLICATION_GRAPH.json#cs_dijkstra
Computer science07

Formal methods / verification

Abstract Interpretation and CEGAR as Distinction Forgetting and Repair

Abstraction forgets distinctions; spurious counterexamples witness harmful forgetting; refinement restores what the proof obligation requires.

STAGE AREADINESS 3
AbstractionWitnessRepair loop
PUBLICATION_GRAPH.json#cs_cegar
Mathematics08

Linear algebra / numerical analysis / PDE

Schur Complements, Static Condensation, and Dirichlet-to-Neumann Reduction

Interior degrees of freedom may be eliminated while retaining the response required at an exposed boundary. Schur reduction becomes an exact calibration of obligation-relative forgetting.

STAGE AREADINESS 5
Exact reductionBoundary responseSource-read priority
PUBLICATION_GRAPH.json#math_schur_dtn
Physics09

Quantum information / operator algebras

Petz Recovery: Recovery Is Not Inversion

Recovery depends on the channel, reference state, and sufficient family. Successful recovery is conditional and typed; it does not make coarse-graining globally invertible.

STAGE AREADINESS 5
RecoverySufficiencySource-read priority
PUBLICATION_GRAPH.json#phys_petz
Mathematics10

Geometric topology / geometric analysis

Poincaré–Perelman as a Boundary Repair Architecture

Ricci flow exposes continuation defects while surgery performs controlled local repair under global proof obligations, without altering the status or content of Perelman’s theorem.

STAGE AREADINESS 5
RepairMonotone witnessProof-status firewall
PUBLICATION_GRAPH.json#math_poincare_perelman
Mathematics11

Numerical analysis / PDE

Galerkin Projection as Obligation-Preserving Reduction

Projection may forget degrees of freedom only relative to the quantity, norm, or proof obligation that must survive. Later obligations can require reopening the representation.

STAGE AREADINESS 5
ProjectionError obligationSource-read priority
PUBLICATION_GRAPH.json#math_galerkin
Physics12

Celestial mechanics / dynamical systems

CR3BP: Hill Regions and Transport Boundaries

Integrals and invariant structures define admissible regions and transport channels in phase space, making boundary language concrete as reachable dynamical geometry.

STAGE AREADINESS 5
TransportReachabilityPhase space
PUBLICATION_GRAPH.json#phys_cr3bp
Physics13

Classical / quantum gauge theory

Gauge Symmetry: Redundant Representation versus Physical Distinction

Different mathematical descriptions may encode the same physical state. Representational difference must not be promoted to physical difference without gauge-invariant content.

STAGE AREADINESS 4
GaugeEquivalenceInvariant content
PUBLICATION_GRAPH.json#phys_gauge
Mathematics14

Topology / algebraic geometry

Sheaves and Gluing as Local-to-Global Closure

Local representations form a global object only when overlap and compatibility obligations close; failed gluing makes the local-to-global defect explicit.

STAGE AREADINESS 4
GluingCompatibilityLocal → global
PUBLICATION_GRAPH.json#math_sheaves

Moves continuously · hover or focus to pause · use the full-height controls to step.

HOW TO READ THESE PUBLICATIONS

A publication should make its status and uncertainty easier to see.

The sections below explain how the Lab separates visibility, maturity, evidence, review, and correction so readers can judge the work without inheriting confidence from presentation alone.

PUBLICATION ≠ PROOF

Publishing something does not make it correct.

Visibility, polish, execution, and release do not automatically increase scientific validity.

PUBLICATION ANATOMY

What a publication page should expose.

Enough surrounding state for a reader to judge the work independently.

STATUS SHOULD BE VISIBLE

The website reports status; it does not invent it.

Where the research or publication record has a more precise status, that source takes precedence.

HELP THE READER JUDGE

Typography should not substitute for evidence.

Readers should be able to distinguish observation, interpretation, conjecture, formal result, empirical support, and deployment evidence.

PUBLICATION TYPES

Different artifacts carry different burdens.

Working papers, reports, notes, specifications, implementations, analyses, and packets are not interchangeable.

CRITIQUE THIS WORK

A useful publication can be corrected.

Substantive criticism, counterexamples, and failed reproductions should have a route back into the research record.

PUBLIC INDEX

Simple to browse; detailed when needed.

The public index should help readers find work without forcing them through the Lab's internal operations system.

STEWARDSHIP

Publication creates long-lived obligations.

Discoverability can outlive assumptions, terminology, evidence, and successor work.

FLAGSHIP RESEARCH-OBJECT PATTERN

A reusable pattern for public research pages.

Important research pages should expose the same basic information: status, claims, evidence, uncertainty, related work, and correction history.

DESIGN POSTURE

Publication pages should look calmer than the claims they contain.

Readable typography and restrained metadata should invite inspection rather than manufacture confidence.

THE PUBLICATION COVENANT

A reader should be able to say:

The publication succeeds when the reader can locate claims, evidence, uncertainty, machinery, critique, and revision.