# Latched Hyperprior

> Latched Hyperprior Hypothesis (LHH) — a sustained vascular contraction engages the smooth-muscle latch-bridge, durably freezing a circuit as a committed hyperprior isolated from global updating; unlatches when its prediction resolves. Use when modeling durable commitments, trauma/PTSD as cemented priors, latch spirals, or the cross-substrate latch (mechanical/bioelectric/immune/sheaf).

- Skill: `plurigrid/latched-hyperprior` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add plurigrid/latched-hyperprior`
- Raw SKILL.md: https://api.skillmd.com/api/skills/plurigrid/latched-hyperprior/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: MIT
- Author: plurigrid (https://skillmd.com/u/plurigrid)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/plurigrid/latched-hyperprior

---


# latched-hyperprior

The **Latched Hyperprior Hypothesis (LHH)** (Johnson 2023): if a vascular contraction is *held long enough* it engages the **latch-bridge** common to smooth muscle — myosin latches to actin and holds tension *without ongoing ATP*. This durably freezes the nearby circuit, isolating it from conscious experience and global updating: a **committed hyperprior**. The latch releases once the prediction it stabilizes is resolved. This is the **−1 / coplay** leg of the vasocomputation timescale triad: the durable *commit*.

## Use When

- Modeling durable commitments / "holding tension in the body" (latches persist seconds → years)
- Trauma / PTSD as a hyperprior cemented against incoming data
- Latch spirals: latch → ↓blood flow → ↓energy → can't release (migraine, cluster headache, neuropathy)
- The cross-substrate view of a latch (below) — the single fixed point read mechanically, bioelectrically, immunologically, sheaf-theoretically

## What *is* a latch? — one fixed point, four substrates

A latch is not four lookalikes; it is **one self-stabilizing fixed point of the body's distributed inference**, read at four levels at once:

| Substrate | Latch = |
|-----------|---------|
| mechanical | latch-bridge: ATP-free myosin–actin hold (Johnson, LHH) |
| **bioelectric** | a hysteretic Vmem attractor pinned by gap-junction feedback; threshold-triggered, instruction-free hold (Levin) |
| **immune** | a walled-off, immune-defended chronic-inflammatory locus (granuloma logic; `neuroimmune-pruning`) |
| computational | a committed hyperprior isolated from global updating (LHH) |
| **sheaf-theoretic** | `H¹ ≠ 0`: a local section that won't glue — gap-junction closure = a lost restriction map |

- **Bioelectric (Levin)**: bistable Vmem holds its own setpoint with no ongoing instruction (cf. the planarian two-head: a stable bioelectric memory surviving regeneration, genome unchanged). Gap junctions are the **restriction maps** of the body's sheaf: open → local sections glue → `H¹=0` → latches cycle; **closed → local section isolated → `H¹≠0`** → a held prior that can't be corrected by neighbors. Pushed to also recruit proliferation, this is Levin's *cancer = cells that bioelectrically disconnect from the morphogenetic field* — a latch that learned to copy itself.
- **Löb fixed point**: `□(commitment) → commitment`. The latch is the contact locus = fixed-point set of the body's `□(self-model)` — the oldies *contact manifold = fixed-point set of □(Nash)* at somatic scale.
- **Release is active**: unlatching costs activation energy (Johnson) = immune **resolution** programs (resolvins/lipoxins) actively terminate, not passive decay. Sauna + cold plunge, attention into the tissue, and psychedelics force the clench–release cycle.

## GF(3) Balanced Triad

```
compressive-vasomotion (+1) ⊗ vascular-clamp (0) ⊗ latched-hyperprior (−1) = 0 (mod 3)
```

**Skill Trit**: −1 (Coplay / commit — the consolidated hold; a *sticky* latch is a **nogood-H¹** to repair, distinct from **content-H¹** legitimately-held disagreement to preserve).

## Honesty markers

Grounded: latch-bridge ATP-free hold; Vmem bistability/hysteresis; gap-junction uncoupling isolating domains; cancer-as-bioelectric-disconnection (Levin); active inflammation-resolution. **Structural correspondence (not asserted biology)**: latch = Löb fixed point = `H¹` of the body's justification sheaf; gap junctions = restriction maps.

## Concomitant Skills

| Skill | Trit | Interface |
|-------|------|-----------|
| `vascular-clamp` | 0 | upstream: sustained clamp → latch |
| `compressive-vasomotion` | +1 | upstream: the sweep |
| `neuroimmune-pruning` | 0 | immune maintenance / unlatch-or-defend |
| `neural-potentiation` | −1 | latch annealed → synaptic prior |
| `sheaf-cohomology` | 0 | H¹ ≠ 0 = local section that won't glue |
| `structural-stability` | −1 | hysteresis / bistable attractor analysis |
| `waddington-landscape` | 0 | canalization of a cemented hyperprior |

## Current literature (2024–2026)

- **Hai & Murphy (1988); Dillon et al. (1981); Rembold (2004)** — the four-state latch; **detachment is rate-limiting** (slow `k7`), high force at low phosphorylation; an "ultraslow" PKC latch holds even longer.
- **Pezzulo, LaPalme, Durant & Levin (2019)** — planarian two-headed form is a **stable bioelectric attractor**, rewritten by gap-junction blockade (octanol), reset by the ion-pump blocker SCH28080 — permanence with a wild-type genome.
- **Levin (2014, Mol Biol Cell)** — Vmem as epigenetic switch; cancer = loss of gap-junction coupling (bioelectric disconnection).
- **van der Kolk (2014); Schleip (fascial plasticity)** — sustained load "freezes" a defensive posture; collagen cross-linking gives the weeks–years tail.
- **Capdeville-Atkinson (1994); FHM2 Ca²⁺-sensitization (2018)** — cluster-headache temporal arteries show spontaneous rhythmic VSMC contraction; MYPT1 = a genetic latch-bias.
- **Key sharpening**: a latch is **kinetic, not thermodynamic** — a fixed point defended by a *rate-limited exit barrier* (hard to *leave*, not hard to *hold*). `MLCK↑ / MLCP↓` is the commit knob; gap-junction uncoupling = isolation from global updating, *literally*.
- **Hook / falsifier**: heat (sauna) and Ca²⁺/Mg²⁺ shifts favour release; octanol flips the attractor and **SCH28080 reset** falsifies "permanent."
- **Grounded**: latch kinetics, Vmem/GJ bistability, fascial cross-linking, VSMC hypercontractility. **Speculative (structural analogy)**: the sheaf-`H¹` / Löb-fixed-point layer — no cohomology is computed; keep as content-H¹ hypothesis, not asserted biology.

## Counterfactual structure

A latch is a **held counterfactual** — a prediction kept against reality until action makes it true (or it is abandoned). This gives LHH a causal-inference reading:

- **Treatment effect**: `latch-above-baseline = E[tension | do(hold)] − E[tension | never]` — the causal effect of the contraction (factual vs counterfactual), already computed in `vasocompute.bb`.
- **Latch spiral = `do(ischemia)` (Pearl rung 3)**: ATP depletion blocks cross-bridge detachment (rigor → `k7→0`), so the latch **cannot** release. Counterfactual harm = `tension(ischemic) − tension(factual)` = "had flow been restored, this tissue *would have* released." The **Mongolian gerbil** (incomplete circle of Willis) is the `do(occlude)` preparation that measures it (CA1 delayed neuronal death).
- **Overhypothesis** (Kemp & Tenenbaum): a latch stores not a datum but a *learned prior over priors* — the computational content of a hyperprior.
- **Resolvability = the repair criterion**: a latch whose counterfactual can be made true is a **nogood-`H¹`** to repair (release on resolution); one in an uncontrollable domain (counterfactual unsatisfiable) is **content-`H¹`** — the irreducible suffering term (TUAI failure mode 2).

See `chirho-counterfactual` for the rigorous SCM version of the `do(k7→0)` query.

## References

- Johnson, M.E. (2023). *Principles of Vasocomputation, Part I*. opentheory.net (§V, LHH; latch-bridge).
- Levin, M. (2022). *Technological Approach to Mind Everywhere (TAME)*. Front. Syst. Neurosci. 16.
- Juliani, Safron, Kanai (2023). *Deep CANALs*. doi:10.31234/osf.io/uxmz6.
- Moore, C.I. & Cao, R. (2008). *The hemo-neural hypothesis*. J. Neurophysiol. 99(5).