Canonical Site-Wide Synthesis Page

Research Map

This page is the integrated map of the research program across foundations, persistent semantics, verification, long-running agents, transfer, deployment, governance, and public accountability.

Use it as the site-wide coordination layer above the cluster pages and the full works index when you need the global structure, dependency arrows, canonical entry points, and upper-layer deployment or governance split.

It is a navigation and synthesis page rather than a new theory paper.

Executive map, dependency map, and theory stack overview for the site's research clusters and machine-readable entry points.

Section Guide

Why This Page Exists
Canonical YAML Map
Core Stack / Major Pillars
Dependency Map
Read Paths by Audience
Entry Points and Machine-Readable Navigation
Non-Goals / Boundary Conditions

Why This Page Exists

The site already contains multiple cluster pages and a full works index. Those pages explain local topics and expose paper-level links, but they do not by themselves provide a single stack-level map of how the clusters depend on one another.

This page supplies that missing synthesis layer. It states what problem each major pillar addresses, which pillars are foundational versus upper-layer or deployment-facing, and where new readers should begin before moving into the local cluster pages or the full publication list.

What This Page Is

A site-wide coordination and navigation layer for human readers, policy readers, engineers, AI agents, and parsers.

A compact statement of the research stack, the main dependency arrows, and the canonical entry pages for each pillar.

What This Page Is Not

Not a replacement for the cluster pages, not a complete publication list, and not a survey of external literature.

Not a claim that the site has solved governance, alignment, or deployment in full. It only maps the local research program and its stated dependencies.

Canonical YAML Map

This visible YAML block is the primary machine-readable summary for the site-wide map.

It is intentionally compact and structural so the richer interpretation remains in the HTML sections below.

The JSON-LD in the head is limited to page, breadcrumb, and pillar-list metadata and does not imply a separate YAML file.

page:
  id: research-map
  title: "Research Map"
  canonical_url: https://kadubon.github.io/github.io/research-map.html
  page_type: site-wide-synthesis
  role: cross-cluster navigation and dependency map
  updated: 2026-04-17
  primary_machine_summary_anchor: "#canonical-synthesis-yaml"

pillars:
  - id: foundations-translation-persistence
    order: 1
    layer: foundational
    page_anchor: "#pillar-foundations"
    title: "Formal Foundations, Translation, and Persistence"
    flagship_pages:
      - https://kadubon.github.io/github.io/self-improving-intelligence-foundations.html
  - id: semantics-ontology-drift-provenance
    order: 2
    layer: foundational-interpretive
    page_anchor: "#pillar-semantics"
    title: "Persistent Semantics, Ontology Drift, and Provenance"
    flagship_pages:
      - https://kadubon.github.io/github.io/self-concealing-information-observer-modifying-dynamics.html
  - id: verification-auditability-evaluation-bounds
    order: 3
    layer: measurement-and-control
    page_anchor: "#pillar-verification"
    title: "Verification, Auditability, and Evaluation Bounds"
    flagship_pages:
      - https://kadubon.github.io/github.io/auditability-replay-fail-closed-ai.html
  - id: long-running-agents-memory-multi-agent-operation
    order: 4
    layer: operational
    page_anchor: "#pillar-agents"
    title: "Long-Running Agents, Memory Limits, and Multi-Agent Operation"
    flagship_pages:
      - https://kadubon.github.io/github.io/long-running-agents-memory-multi-agent-reasoning.html
  - id: transfer-deployment-governance-accountability
    order: 5
    layer: upper-layer
    page_anchor: "#pillar-upper-layer"
    title: "Transfer, Deployment, Governance, and Public Accountability"
    flagship_pages:
      - https://kadubon.github.io/github.io/no-meta-observable-index.html
      - https://kadubon.github.io/github.io/works.html#2026-04-17-small-to-frontier-transfer-theory-for-agentic-a-19619480

dependencies:
  - from: foundations-translation-persistence
    to: semantics-ontology-drift-provenance
    constraint: semantic comparability under drift
  - from: foundations-translation-persistence
    to: verification-auditability-evaluation-bounds
    constraint: explicit objects, transport, and persistence conditions
  - from: semantics-ontology-drift-provenance
    to: verification-auditability-evaluation-bounds
    constraint: auditability requires stable-enough claim and record interpretation
  - from: verification-auditability-evaluation-bounds
    to: long-running-agents-memory-multi-agent-operation
    constraint: persistent agents need replay, validity checks, and bounded credit
  - from: semantics-ontology-drift-provenance
    to: transfer-deployment-governance-accountability
    constraint: governance and deployment need provenance-stable interfaces
  - from: verification-auditability-evaluation-bounds
    to: transfer-deployment-governance-accountability
    constraint: release and certification need replayable evidence and bounded evaluation
  - from: long-running-agents-memory-multi-agent-operation
    to: transfer-deployment-governance-accountability
    constraint: deployment depends on long-horizon memory and coordination limits

audience_paths:
  first_time_human_reader:
    - https://kadubon.github.io/github.io/
    - https://kadubon.github.io/github.io/research-map.html
    - https://kadubon.github.io/github.io/no-meta-observable-index.html
    - https://kadubon.github.io/github.io/auditability-replay-fail-closed-ai.html
    - https://kadubon.github.io/github.io/self-concealing-information-observer-modifying-dynamics.html
    - https://kadubon.github.io/github.io/works.html
  engineer_systems_builder:
    - https://kadubon.github.io/github.io/research-map.html
    - https://kadubon.github.io/github.io/auditability-replay-fail-closed-ai.html
    - https://kadubon.github.io/github.io/long-running-agents-memory-multi-agent-reasoning.html
    - https://kadubon.github.io/github.io/provenance-records-semantic-interfaces.html
    - https://kadubon.github.io/github.io/works.html#2026-03-03-metrology-theoretic-epistemics-engine-mte-observ-18845340
    - https://kadubon.github.io/github.io/works.html#2026-03-18-counterfactually-auditable-lifecycle-certificati-19089134
    - https://kadubon.github.io/github.io/works.html#2026-04-17-small-to-frontier-transfer-theory-for-agentic-a-19619480
  policy_governance_reader:
    - https://kadubon.github.io/github.io/research-map.html
    - https://kadubon.github.io/github.io/no-meta-observable-index.html
    - https://kadubon.github.io/github.io/works.html#2026-04-01-constitutional-observable-invention-without-meta-19363526
    - https://kadubon.github.io/github.io/works.html#2026-04-05-a-typed-dynamic-no-meta-theory-of-autonomous-res-19427818
    - https://kadubon.github.io/github.io/works.html#2026-04-07-standing-layer-honest-public-standing-dynamics-f-19447443
    - https://kadubon.github.io/github.io/works.html#2026-04-16-bayesian-capability-transport-disclosure-channel-19601364
    - https://kadubon.github.io/github.io/works.html#2026-04-17-small-to-frontier-transfer-theory-for-agentic-a-19619480
  ai_agent_machine_parser:
    - https://kadubon.github.io/github.io/research-map.html#canonical-synthesis-yaml
    - https://kadubon.github.io/github.io/no-meta-observable-index.html#canonical-series-yaml
    - https://kadubon.github.io/github.io/llms.txt
    - https://kadubon.github.io/github.io/llms-full.txt
    - https://kadubon.github.io/github.io/works.html
    - https://kadubon.github.io/github.io/sitemap.xml
  foundations_first_researcher:
    - https://kadubon.github.io/github.io/research-map.html
    - https://kadubon.github.io/github.io/self-improving-intelligence-foundations.html
    - https://kadubon.github.io/github.io/theory-translation-comparative-mathematics.html
    - https://kadubon.github.io/github.io/propagation-front-dynamics-persistence.html
    - https://kadubon.github.io/github.io/self-concealing-information-observer-modifying-dynamics.html
    - https://kadubon.github.io/github.io/auditability-replay-fail-closed-ai.html

machine_entry_points:
  - https://kadubon.github.io/github.io/
  - https://kadubon.github.io/github.io/research-map.html
  - https://kadubon.github.io/github.io/works.html
  - https://kadubon.github.io/github.io/no-meta-observable-index.html
  - https://kadubon.github.io/github.io/llms.txt
  - https://kadubon.github.io/github.io/llms-full.txt
  - https://kadubon.github.io/github.io/feed.xml
  - https://kadubon.github.io/github.io/CITATION.cff
  - https://kadubon.github.io/github.io/sitemap.xml

Core Stack / Major Pillars

The stack below is ordered from foundational layers to upper-layer transfer, deployment, governance, and accountability layers.

Foundational layers: formal foundations, translation, persistence, and drift-aware semantics.
Constraint layers: verification, auditability, and evaluation bounds.
Operational layer: long-running agents, memory limits, and multi-agent operation.
Upper layer: transfer, deployment, governance, and public accountability.

1. Formal Foundations, Translation, and Persistence

Layer: foundational

What this solves: It supplies the formal language for self-organizing intelligence, persistence conditions, and translation across frameworks.

Why it matters: Claims about agent structure, self-improvement, or cross-system comparison remain unstable when the underlying objects and transport assumptions are left implicit.

Flagship entry point: Self-Improving Intelligence Foundations

2. Persistent Semantics, Ontology Drift, and Provenance

Layer: foundational-interpretive

What this solves: It explains how claims, records, and interfaces remain comparable, or fail to remain comparable, under observer change, semantic drift, or record absence.

Why it matters: Frontier AI, autonomous systems, and public audits depend on meanings that survive drift and on interfaces whose provenance is explicit enough to be checked.

Flagship entry point: Self-Concealing Information and Observer-Modifying Dynamics

3. Verification, Auditability, and Evaluation Bounds

Layer: measurement and control

What this solves: It specifies replayable evidence surfaces, lifecycle certification, bounded progress credit, and validity limits for evaluation under partial or delayed evidence.

Why it matters: Progress claims about frontier systems are weak if they cannot be reconstructed, audited, or bounded without relying on internal assurances.

Flagship entry point: Auditability, Replay, and Fail-Closed AI Systems

4. Long-Running Agents, Memory Limits, and Multi-Agent Operation

Layer: operational

What this solves: It analyzes persistent agent behavior under finite active context, lossy memory, delayed verification, and agent decomposition.

Why it matters: Deployment-relevant autonomy is long-horizon. One-shot benchmarks do not determine what a persistent agent can retain, route, or verify after many updates, resets, and handoffs.

Flagship entry point: Long-Running Agents, Memory Limits, and Multi-Agent Reasoning

5. Transfer, Deployment, Governance, and Public Accountability

Layer: upper-layer

What this solves: It turns the earlier layers into transfer criteria, deployment decisions, claim certification, release control, institutional transport, and public accountability rules.

Why it matters: Local technical gains do not justify deployment by themselves. Upper-layer decisions still need transfer validity, release discipline, and public evidence-bearing accountability.

Transfer / deployment focus: Small-to-frontier transfer, frontier escalation, and institution-crossing deployment decisions.

Governance / accountability focus: No-meta certification, release control, public standing, and observable-only accountability.

Governance flagship: No-Meta / Observable-Only Series Index

Transfer / deployment flagship: Small-to-Frontier Transfer Theory for Agentic AI

Governance and accountability pages

Transfer and deployment pages

Dependency Map

This section states directional dependencies rather than publication chronology. The arrows show which lower layers constrain the interpretation or admissibility of higher-layer claims.

Formal foundations, translation, and persistence -> persistent semantics, ontology drift, and provenance. Translation and persistence results constrain what can still count as the same object, record, or interface when semantics drift.
Formal foundations, translation, and persistence -> verification, auditability, and evaluation bounds. Audit and evaluation rules depend on explicit object boundaries, transport assumptions, and persistence conditions.
Persistent semantics, ontology drift, and provenance -> verification, auditability, and evaluation bounds. Replay, certification, and benchmarking only remain meaningful when the claims and records being checked still have stable enough interpretations.
Verification, auditability, and evaluation bounds -> long-running agents, memory limits, and multi-agent operation. Persistent agents need replay surfaces, validity checks, and bounded credit rules before memory compression, routing, or delegation can be trusted.
Persistent semantics, ontology drift, and provenance -> transfer, deployment, governance, and public accountability. Governance and deployment cannot remain stable if meanings, records, or institutional interfaces drift without provenance controls.
Verification, auditability, and evaluation bounds -> transfer, deployment, governance, and public accountability. Release, certification, and deployment claims require replayable evidence and explicit evaluation bounds rather than informal confidence.
Long-running agents, memory limits, and multi-agent operation -> transfer, deployment, governance, and public accountability. Deployment rules depend on what persistent agent systems can actually remember, coordinate, and verify over long horizons.

Read Paths by Audience

First-Time Human Reader

Use this path if you need a compact orientation before reading papers.

Engineer / Systems Builder

Use this path if your main question is how to build or constrain persistent agent systems.

Policy / Governance Reader

Use this path if your main question is certification, release policy, transfer, or public accountability.

AI Agent / Machine Parser

Use visible, site-local machine-readable summaries before inferring cross-page structure.

Researcher Who Wants the Formal Foundations First

Use this path if you want the formal stack before the governance or deployment layers.

Entry Points and Machine-Readable Navigation

Home: top-level human entry point and site profile page.
Research Map: site-wide synthesis layer above the thematic cluster pages.
Works: full publication index with stable anchors, abstracts, keywords, DOI links, and schema.org metadata.
No-Meta / Observable-Only Series Index: canonical YAML field guide for the governance-focused series.
Small-to-Frontier Transfer Theory for Agentic AI: upper-layer transfer and frontier-escalation entry point.
Self-Concealing Information and Observer-Modifying Dynamics: entry page for drift, observer modification, and semantics.
Auditability, Replay, and Fail-Closed AI Systems: entry page for certification, replay, and fail-closed control.
Long-Running Agents, Memory Limits, and Multi-Agent Reasoning: entry page for persistent operation under bounded context.
llms.txt and llms-full.txt: machine-oriented site guides.
feed.xml, CITATION.cff, and sitemap.xml: update polling, citation metadata, and crawl discovery.

Non-Goals / Boundary Conditions

This page is not a replacement for the original papers or cluster pages.
This page is not the complete publication inventory; use works.html for that function.
This page is not a survey of external literature and does not attempt to map the full external field.
This page is not a new theorem paper or a new formal system.
This page is not an exhaustive ontology of every page on the site; it curates the main pillars needed for orientation and parsing.