PVM V6 · the visual runtime and the instrument taxonomy
PVM V6 · the visual runtime and the instrument taxonomy
A handoff note. The agentprivacy.ai/model page gained an interactive visual
runtime ("the Cartographer"). This note records the findings as proposed
additions to V6, so a future agentprivacy-docs context can formalize them. The
register at head C89 remains the authority; nothing here re-numbers it.
Why this note exists
The /model page is the math-atomic surface of the framework. It read as an
open path; it now also walks as one. The build (see the companion chronicle,chronicles/2026-06-28_v6_visual_runtime_instruments.md) added a small
data-driven runtime that binds one interactive instrument per term or
conjecture and pulls each claim's honesty (confidence · status · home) live
from the register mirror. The seed for the design wasresearch/limitative-theorems-and-privacy-is-value.md and its companionlimitative-theorems-explorer.html.
The runtime surfaced three things worth folding into V6.
1. The instrument taxonomy (a reusable unit, not a per-conjecture gallery)
The register does not need one bespoke visual per claim. It needs a small set of
reusable instrument kinds; each conjecture is a configuration of a kind.
Seven kinds cover most of the register:
| kind | what it teaches | register targets (examples) |
|---|---|---|
| trace | two chains joined across a ⊥ seam; click a correspondence | C85 ARCH-1 bridge · the Gödel/Tarski mapping |
| gate | a multiplicative product collapsing when any factor → 0 | C7 three-axis (the falsification frontier) · whole V(π,t) |
| curve | a value falling/rising against a parameter or time | C81 existence-leak · C82 moving ceiling · e^(−λt) · C30–C33 half-life |
| dial / inversion | one parameter driving parallel meters to a shared collapse | C17 amnesia > policy · the completeness inversion |
| lattice / geometry | the 64-vertex lattice, strata, bnot-pairs, the stella octangula | C14 dihedral · C53–C54 bnot-pairs · C88 parity cube · C89 octahedral gap |
| phase | a trajectory through state space | C18–C21 Lorenz · T_∫(π) path integral |
| fold / compose | accumulation and the exponential-to-linear gap | C83 compositional leakage · C87 Key-as-accumulator (IVC) |
(Architectural. The taxonomy is the unit of build work, not the conjecture
count. Observation-only / reserved rows — e.g. C41, C57, C62 — get no
instrument; a visual earns its place only where a parameter can be moved or a
structure traced.)
2. The binding rule — honesty is read, never embedded
Every instrument binds to a register id and reads its honesty live. The
visual owns its interaction and its illustrative model; the register owns the
claim. A curve never states a confidence — it asks the register for one. If the
register moves C81 from 70% to a higher stage, the existence-leak frame moves
with it, with no edit to the visual. This makes the existing rule ("when a
surface and the register disagree, the register wins") mechanical rather
than editorial.
A load-bearing field rides on each instrument: model_label ∈ {illustrative, structural, proven}. It forces every shape to declare whether it is the proven
relation or a chosen teaching model, and the frame marks it visibly. The
existence-leak curve's convex shape is illustrative; the three-axis gate's
multiplicativity is structural.
(Operational. Implemented as a manifest + a frame that calls the register
mirror at render time.)
3. The drift correction (a concrete V6 erratum the runtime forced)
The seed explorer hard-coded its confidences. It rendered the existence-leak
curve at ~55% and labelled it ~C40 — both inherited from a stale source.
The live register is C81 at ~70% ("Existence-Leak: a ZK proof of
feasibility leaks an upper bound on reconstruction difficulty; I(feasibility;
method) > 0", PROMOTED Run 3 2026-06-10). The runtime binds the curve to C81
and the frame reads ~70% live.
Proposed erratum for any prose still carrying the seed's numbers: the
existence-leak instrument is C81 (~70%), not C40 (55%). C40 is the Zcash
dual-ledger entry in the Bound Collection; the generic "~C40" the limitative
note used for existence-leak should be read as C81 going forward. The
limitative note (research/limitative-theorems-and-privacy-is-value.md) already
touches C7, C14, C81, C85, C86 — this just fixes the existence-leak pointer.
(Operational — the drift was observed directly; the correction is a pointer
fix, not a re-numbering.)
4. The two-register tie — the Cartographer ⊥ the Chart Shop
The City already narrates this runtime in advance. The Navigation District at
V44 — the Chart Shop, Pleione's harbour — is described as an
attentional shop holding pre-episodic constellations in suspension until the
bearer chooses a release-direction (astrolabe + the Φ-gap-at-epistemic-register
as primitives). That is a visualisation runtime narrated before it was built.
So the operational runtime is the Chart Shop's math-atomic twin. Proposed
naming: the Cartographer (the runtime) renders charts (the instruments),
narrated by the Chart Shop at V44. By the convergence rule, when a math term the
Chart Shop narrates gains a concrete movable instance, confidence rises on both
sides. No such tie currently exists in the docs — this note opens it.
(Anticipated. The narrative anchor exists; this names the operational instance
against it. A future context may want a city-register conjecture for the
Cartographer ⊥ Chart Shop convergence — candidate, not registered here.)
5. Geometry is not greenfield (deferred, but mapped)
The geometry/phase instruments already exist as prototypes: soulbis.com/star
(in C:\Users\mitch\star) renders the stella octangula manifold with a live
lattice console (ε separation depth, det(Σ) core scale, the ⚔️:🧙 ratio, n = 6
azimuthal bits, the ∂M 96/64 boundary) and a City Key import that walks a
real artifact's path. The /lattice, /sigil, /skye siblings are the same
family. Phase 3 of the runtime is absorption + reconciliation (V5.4 → V6) +
register binding (ε/det(Σ)/ratio → Φ_agent/C1; 96/64 → C6; neg·bnot·succ →
C14; the star tetrahedron → C88/C89; the path walk → T_∫/C18–C21), not first
build. The City Key import should generalize to a first-class artifact-input
instrument parameter (a curve over a real disclosure history, a gate over a real
provider set, a lattice walking a real key).
(The star is Operational, observed at soulbis.com/star. Binding to V6 is
Anticipated.)
Proposed V6 additions — summary for the next context
- Adopt the seven-kind instrument taxonomy as the canonical visual
vocabulary of V6 (trace · gate · curve · dial · lattice · phase · fold). - Record the binding rule (honesty read-live,
model_labelmandatory) as a
V6 surface-discipline alongside "the register wins." - File the drift erratum: existence-leak = C81 (~70%), retiring the
seed'sC40/55%. - Open the Cartographer ⊥ Chart Shop (V44) convergence as a candidate
city-register tie (not registered here; flagged for disposition). - Map the star/lattice absorption as the Phase-3 geometry binding (no new
conjectures; reconciliation + binding of C6, C14, C88, C89, Φ_agent, C1).
No conjecture numbers are claimed or moved by this note. Register head stays
C89; next free C90.
Verify: agentprivacy.ai/model · agentprivacy.ai/model/atlas ·agentprivacy_master/src/lib/visuals.ts · src/components/model/visuals/
(⚔️⊥⿻⊥🧙)😊