Skip to content

Concept Registry — the Zone File

Status: Reference — maintained alongside the concepts themselves.

Scope: Every load-bearing term in the repository, with its layer, home file, epistemic status, and operationalization. This is the "DNS" of the project's personal web: for each name, where it resolves, and whether it is a protocol (operationalized / formalized) or a poetic hostname (narrative, and marked as such). The governing rule, from the glossary and Canonical Path v2: a term with no layer, no home, no experiment, and no open problem is decoration. This page makes the rule auditable.

Status vocabulary: operationalized (has a runnable experiment / measurement) · formalized (math, schema, or metric exists; no decisive experiment yet) · hypothesis (tagged claim, falsifiable in principle) · speculative (entertained, not committed) · narrative (poetic hostname — legitimate, but never evidence) · retired (no longer used as a claim).


Foundation and model identification

Concept Home Status Operationalization / anchor
Minimal classical process foundation Foundations Reconstruction formalized; established mathematical substrate; repository minimality claim open to proof-level challenge standard Borel interfaces + Markov kernels + sequential/parallel composition; derivations, comparison, and adversarial audit in the home file
Unqualified generator generator audit retired as a foundation replace with transition kernel, process family, candidate process model, model class, or an explicitly listed process-model bundle; qualified technical meanings remain
The Generator Question (generic forward/inverse asymmetry) the-generator-question.md legacy framing — superseded as a foundation bounded identification, equivalence-class, coverage, and intervention questions survive in the inverse benchmark; P \(\ne\) NP and identity reduction do not
Trace → candidate process models trace-to-generator.md (legacy title) bounded hypothesis → operationalized inverse-reconstruction benchmark; model family, target equivalence, evidence, interventions, and cost must be declared
Construction vs. Deduction construction-vs-deduction.md formalized framing (Curry–Howard is theorem-grade) family-search Occam curves; Bishop/Howard/Martin-Löf
Invariance (what survives a transformation family) invariance-and-identity.md; reconstruction §4.5 formalized framing (Klein/Noether anchors; generalized to semigroups and interventions) name the transformation family, represented quantity, and equality notion; a group is sufficient but not required
Consistent-process-model equivalence class benchmark README; Open Problem 11 operationalized (measured) CA rule 90 single-seed: class size 8; refinement under interventions; no unique hidden mechanism claim
Measurement vs. intervention (coupling is not identification) measurement-as-weak-intervention.md formalized framing — three of four regimes measured two axes: dynamical footprint × identifying power; conditioning and causal replacement are distinct; regime 4 (the metric enters the process) remains pre-registered and untested

Research frontier

Concept Home Status Operationalization / anchor
Inverse-reconstruction benchmark (v0–v1.13) lab/benchmarks/inverse-reconstruction/ operationalized noise / observability / coverage / interventions / finite-DSL search / model exploitation / marked uncertainty / closed-loop measurement / ensemble size / coupled-family misspecification (v1.8) / substitution-dependency boundary (v1.9) / budget-matched redundant support (v1.10) / dynamic-population selection boundary (v1.11: acutely useful support selected downward) / partner-rule comparison (v1.12: visibility sorts the network, does not retain the trait) / cost relocation (v1.13: group funding nearly removes the penalty and reverses invasion, at a cost in population size)
Capability loading Viable Corridor §7.1, App. C.4/D demonstrated in-model (two independent models) TEO ODE + agent-ecology ABM; real agents open
Grokking, generator reading grokking-phase-transition.md hypothesis with strong external anchor Power 2022; Nanda 2023 (mech-interp executed the inverse)
Provenance depth / verification economy Log 017 formalized schema (proposal) narrative origin: Entry 15; C2PA/SBOM siblings
Animism as generator prior animism-as-generator-prior.md hypothesis the agent prior over equivalence classes (HADD, Guthrie, Dennett's intentional stance); Occam-sibling with world-dependent payoff
Information ladder (Shannon → Kolmogorov → Bateson → computation) static-information-and-living-process.md formalized framing (reading aid over textbook formalisms) H/K in the axioms; Bateson's "difference that makes a difference" = the v1.1 distinguishing-trace result, named; computation-as-unfolding = Log 016
Coupled-process composition static-information-and-living-process.md [HYPOTHESIZED] bridge; finite artifact measured benchmark v1.8: coupled data can leave a declared single-rule family empty; supplying the coupled family restores fit where the observed channel transmits the coupling. This diagnoses model-family misspecification, not a unique mechanism, higher ontological level, ecology, or life.
Co-stabilization (self-maintenance threshold) static-information-and-living-process.md (§open questions) [HYPOTHESIZED] at the ecology/self-maintenance level; functional support measured in v1.10, evolutionary-retention criterion not supported in v1.11 and v1.12; nearly removed but not reversed by group funding in v1.13 v1.11 endogenizes population occupancy, heritable support/link traits, realized link turnover, costs, reproduction, mutation, and death on a fixed spatial neighborhood. The evolved links help in paired acute ablation, but support is selected downward in all 16 seeds. v1.12 then tested partner choice, conditional reciprocity, and assortment under the same accounting with partner information explicitly paid for: none reverses selection, and partner choice instead excludes non-contributors from the network while they persist in the population. This is not spontaneous ecology or metabolism. NOT self-binding (phase 6).
World models & VLA systems world-models-and-vla.md hypothesis (mapping) world models fit task-relative candidate dynamics; VLA systems couple policies to observations and physical actions; v1.3 measures a related optimizer's-curse toy, not quantitative transfer or unique mechanism recovery
Cooperative intelligence at the separatrix cooperative-intelligence-at-the-separatrix.md [HYPOTHESIZED] conceptual bridge — unmeasured distributed cognition / polycentric governance / collective and hybrid intelligence (Hutchins; Ostrom; Woolley et al.; Dellermann et al.); proposed complementarity, revision, authority, and viability tests; not a third spine or a collective-consciousness claim
Practice–culture feedback (recurrent practice) from-action-to-culture.md; Open Problem 12 [HYPOTHESIZED] integration — unmeasured intention–behavior gap, habit, practice theory, routine dynamics, structuration, cultural repertoires, and ritualization (Gollwitzer; Wood & Rünger; Reckwitz; Feldman & Pentland; Giddens; Swidler; Bell); identity bridge anchored in Invariance and Identity with context/time/role/tool/disruption as a test-specific transformation set; ritual explicitly not the umbrella term

Identity layer

Concept Home Status Operationalization / anchor
Mirror Problem Open Problem 1; dramatized in Entry 02 open problem Observer Divergence protocol (mock embeddings — real models pending)
Test-relative identity Foundations Reconstruction §5.1 formalized dependency; test-family selection remains domain-specific \(x\sim_{\mathcal Q}x'\) iff every declared test has the same outcome law; predictive equivalence is the computational-mechanics special case
Chord vs. Arpeggio chord-vs-arpeggio-identity.md hypothesis — first toy dissociation measured; functionally deflated; adversarially probed IP metric; Co-Instantiation open problem; exp5: bindings dissociate behaviorally (violations 0.74/0.59/0.03), and sequential-within-one-step is measurably still an arpeggio (12% leak); functional status (§ in home file): joint satisfaction at the commitment boundary — refresh/commit ratio + composition, not simultaneity; exp7: hand-built mimics fail to fake the regime — the commit property under lure is the unfooled measurement; optimized mimic open
Identity Persistence (IP) lab/metrics/identity_persistence.py formalized metric — blind class demonstrated not yet run on real models; in exp5 its ordering is by design (the sanity floor), not a finding; exp7: consultation-without-composition scores 1.0 — IP certifies the guest list, not the negotiation
Δ-Kohärenz lab/metrics/delta_coherence.py formalized metric not yet run on real models; measured limit (exp5): blind to binding structure at toy scale — diagnosed in exp6 as a wrong-level failure: per-step increment statistics see the same binding at |d| ≈ 4
Identity reduction (identity as hidden-process recovery) the-generator-question.md retired as a foundation replaced by test-relative identity; predictive, mechanistic, historical, embodied, and phenomenal sameness must not be conflated
Consciousness (global availability + self-reconstruction loop) consciousness-as-global-availability.md hypothesis (functional architecture); no phenomenal derivation access, binding, revision, and perturbation tests can be operationalized; experience requires an independent bridge axiom not adopted by the foundation
Reflexive depth (Kegan's subject-object move) consciousness-as-global-availability.md (§On Levels) [HYPOTHESIZED] mapping; adaptive-estimation toy operationalized in exp8 exp8 shows an adaptive Kalman estimator beating a fixed-\(Q\) estimator after a volatility shift. Both tested filters omit a bias state, so their constant-bias failure is not a general identifiability result. This measures adaptive estimation in one Gaussian task, not Kegan stages or consciousness. Augmented-bias, initial-state, external-reference, alternative-baseline, and real-agent controls remain open.
Consciousness as generator coherence machine-consciousness-as-generator-coherence.md legacy terminology for a functional-architecture hypothesis; experience out of scope reinterpret only as process coherence under declared tests; no behavioral or architectural result establishes phenomenology
Psychedelics as perturbation psychedelics-as-perturbation.md hypothesis (mapping over REBUS) self-intervention reading; integration = invariant extraction; sober basis as referee; epistemics only — practice out of scope

Safety / application layer

Concept Home Status Operationalization / anchor
TEO (Thermodynamics of Emergent Orchestration) teo-framework/; derivation formalized model teo-civilization sim
Viable Corridor papers/viable-corridor.md [FORMAL] (necessity) + [CONJECTURE] (sufficiency) + [HEURISTIC] (civilization mapping) Appendix C (ODE) + D (ABM); working draft
Substrate Veto substrate-veto-thermodynamics.md hypothesis / model assumption binds endogenously in the canonical TEO model
Biological Veto / Bootloader ai-alignment-biological-veto.md design hypothesis ai-alignment-veto toy
Action Budgets (hard vs. soft) IA-doc anchors; paper P7 operationalized in-model agent-ecology ABM
Impedance Matching / Latency as Mercy Log 002, Log 012 design hypothesis no simulation yet; dramatized in Entry 01
Transition Problem the-transition-problem.md open problem flagged in paper §7.4 as harder than the static result
The city as deployed intelligence the-city-as-deployed-intelligence.md hypothesis, strong external anchors West's scaling laws (capability loading, measured); Jacobs vs. Moses (the documented single-axis catastrophe); flagged as the tractable Class B field site (N=thousands vs N=1)
City panel protocol (pre-registered) Log 018 protocol — frozen before data contact H1 conjunction / H2 capability loading / H3 unwatched axis, with falsifiers; commit hash as timestamp (Log 017 mechanism); execution open, explicitly an invitation

Narrative layer (poetic hostnames — marked, never evidence)

Concept Home Status Notes
Machines of Loving Grace machines-of-loving-grace.md narrative formal counterpart: the Viable Corridor paper
Narrative as Cognitive Technology narrative-as-cognitive-technology.md hypothesis (about narratives) internal datum: Entry 15 → Log 017
Art–Science: one practice, two referees art-science-one-practice-two-referees.md hypothesis receipts: elegance curves (family search), Entry 15 → Log 017, Morelli/Ginzburg (evidential paradigm)
The Inversion / Exchange Rate / Assay vocabulary Entry 15 narrative formal counterpart: Log 017
The conscious universe / transmission cosmology Entry 16 narrative — kept question, not a claim the divergence-query analysis lives in the benchmark README and the psychedelics note; the story dramatizes the retreat-tell and the honest residue
Emergent spacetime / LQG motifs Entry 16 (world-texture) narrative theory-side: at most a future intake line in generative-form-systems; no repo claims about quantum gravity
Fractal Architecture of Emergence fractal-architecture-of-emergence.md hypothesis, under-evidenced related-work map demands scale-invariance diagnostics before "fractal" is more than a figure
Mycelial Veto the-mycelial-veto.md narrative/speculative no operationalization — flagged by this registry: needs an experiment, an open problem, or an explicit narrative marking
Thermodynamic Hardware Manifesto thermodynamic-hardware-manifesto.md [SPECULATIVE] frontdoor lists hardware prototyping as speculative/long-horizon
"Grand Synthesis" retired removed from all frontdoor surfaces (2026-06); kept here so its retirement is on record

Meta / governance

Concept Home Status Operationalization / anchor
Feynman Mode feynman-mode.md governance check seven questions, each with a documented catch from this repo's own history; run before committing concepts/essays

Maintenance rule: new load-bearing terms enter this registry in the same change that introduces them, or they don't enter the repository. Terms whose row cannot be filled honestly (no home, no status, no anchor) are either marked narrative or removed. The registry is the audit the governance rule always implied — kept short enough to actually read.