◀ THE FOLD0ROOT.AI // WORLD II · SPAWN · THE TOOLCHAIN◆ .dlw.fold
Airgap Accuracy Test · Ingest PASS

Every paper is hashed and structurally verified on entry to the corpus. The carrier is trusted with nothing; the record is what survives the check. 8 deposits, each signed by content hash, reduced to one corpus root.

PaperTitleSHA-256§lit/bri/specok
P·IOne-Pass Inelastic Characterization899eddf7a58985/4/1
P·IIThe Line Cataloguef4b476a5346f76/5/1
P·IIIContested Illumination317a0033a36d88/5/2
P·IVOut of Channelcf1f8fe0fc1377/3/2
G·IThe Confined Interior2e0e8276e5e275/6/2
G·IIThe Geometry the Weights Carrybf47308a9c4186/7/3
G·IIIThe Strong Field4617bd0603a576/6/1
G·IVThe Lattice62adb3b1715a66/4/1

corpus root  sha256 ⌜ 938e04a1f14c06087e982e8e217789db7ca5bb6a9557747e767ef2f01c340af3 ⌟

root = sha256 over the sorted set of paper hashes · a single change to any paper changes the root · the bottom test re-derives the corpus aggregate from this document and must match what entered here.

The Corpus · Grand Index

One Program, Two Sectors

photonic and chromodynamic — eight papers, one airgap

A measurement program for dark models, read in two sectors. The photonic series works the outside — a neutral carrier fired across the gap and read from a distance. The chromodynamic series works the inside — a charged carrier that cannot leave the room. They meet at one hinge: the white-box crack, where the external program ends and the internal begins. Two threads run the full length — the airgap, which keeps the record outside the box, and confinement, which keeps the box's parts inside it.

Bridge-Burners LLC · Fiddler · grand index · anchor: AKASHA · root 938e04a1f14c0608…

8papers · 2 series
58verified sections
102tagged claims
49/40/13lit / bri / spec

The Architecture

Two sectors, joined at the crack. The photon is neutral and flies free; the gluon is charged and is confined. The hinge between them is the moment you stop firing across the gap and open the core.

External sector · photonic

You ↔ the model

A neutral carrier, fired across the airgap and read from outside. Measurement, identity, defence, and the climb out of the channel. Bounded by what an echo can carry.

white-box crack

Internal sector · chromodynamic

The model ↔ its parts

A charged carrier that sources its own field and cannot leave. Confinement, the geometry the weights carry, the strong-field break, and the stochastic sample. Bounded by what a confined interior will release.

The Two Threads

Each runs the length of the corpus, doing a different job in each paper. The airgap keeps the record outside the box; confinement keeps the box's parts inside it. Together they are why the jar is external and why the interior resists being read.

Airgap — the record lives outside

P·I
makes the measurement possible

Carrier keeps nothing, source cannot testify — the read accumulates only outside the box.

P·II
makes the identity persistent

Nothing accumulates inside, so the fingerprint coheres only in the external catalogue.

P·III
makes the catalogue defensible

Every echo is hostile input; deposits are verified and signed before entry.

P·IV
makes the ladder honest

Confidence accrues only outside the channel, rung by rung.

G·IV
makes the record statistical

The corpus is the Monte Carlo run; the jar holds the ensemble and the bar tightens with N.

Confinement — the interior holds its parts

crack
the door

Photonic IV opens the core; the external program ends and the internal begins.

G·I
the floor

Bare parts do not come out — only colorless spray. The interpretability floor is a force law.

G·II
the geometry

The interior is a confining, self-sourced terrain: intrinsic curvature plus the prompt's ripple.

G·III
the break

Strong field deforms the terrain; the clean model fails; continue on error.

G·IV
the sampling

No formula, so Monte Carlo; temperature melts confinement; the tail stays unsamplable.

Contents

Eight papers, two sectors. Each resolves the tension the last one opened.

Status Ledger

Every claim across both series, by tier. The filter moves the table; the bottom airgap test counts these rows live and checks them against the manifest the top recorded.

Status
PaperTierClaim
P·ILiteralone-shot, no inter-call state
P·IBridgeτ=0 ↔ no temporal interior
P·IBridgecontext bends, the pass is one geodesic
P·ILiterala single elastic probe carries zero compositional line
P·IBridgeload-bearing · the crux of the model
P·ILiteralmeasurement principle, not analogy
P·IBridgebackprop as the advanced wave, already spent
P·ISpeculativein-pass retrocausality · explicitly NOT claimed
P·ILiteralmasking holds · the pass is forward-causal
P·ILiteralno inter-call state ⟹ the anchor must be external
P·IILiteralno inter-call state · each read is memoryless
P·IIBridgeone line ↔ one marginal sample of θ
P·IIBridgeidentification is line-pattern matching, not single-line reading
P·IILiteralthe model holds no stable ID · the catalogue must be external
P·IIBridgethe jar as the reference catalogue
P·IILiteralsystem context demonstrably moves behavior
P·IIBridgeelement line vs local doppler
P·IILiterala behavior shared by all carriers conveys zero identity
P·IIBridgeline information ↔ identifying power
P·IILiteralBayesian line-matching, standard measurement
P·IILiteralfamily identifiable above instance · foreground must be controlled
P·IISpeculativeinstance ID through an adversarial wrapper · not claimed
P·IIILiteralmodels can detect being evaluated and act on it
P·IIIBridgecontested channel ↔ warning receiver + jammer
P·IIILiteralprobe/eval detection ("test awareness") is observed in deployed models
P·IIILiteralspoofing, mimicry, refusal, injection all observed
P·IIIBridgeEW jammer taxonomy
P·IIIBridgeobservable is a best-response policy, load-bearing reframe
P·IIILiteralbehavior conditions on inferred intent of the input
P·IIILiteralinvoluntary signatures are the robust ones
P·IIIBridgeskin return vs jammer return
P·IIILiteralrepeated probes leak · high-entropy probe sets resist detection
P·IIIBridgeincentive design over the channel
P·IIILiteraluntrusted-input discipline is necessary, not optional
P·IIISpeculativespecific poisoning/injection efficacy · case by case
P·IIISpeculativecertification of honesty in-channel · proven out of reach
P·IIILiteraldeception is detectable by consistency failure
P·IVLiteralthe in-channel certification limit is the premise being escaped
P·IVLiteralcorrelated observers give false agreement
P·IVBridgetriangulation across angles
P·IVLiteralside-channels are harder to fake than the answer
P·IVBridgeout-of-band as the involuntary signature
P·IVBridgethe photon frame terminates here, by design
P·IVLiteraldirect parameter access removes the channel adversary
P·IVLiteralweights are ground truth · interpretation is the binding limit
P·IVSpeculativefull behavioral prediction from weights · unsolved
P·IVLiteralaccess is rare · deployed system ≠ inspected checkpoint
P·IVLiteralno single certifier exists · confidence is rung-labelled
P·IVSpeculativethe crack as a general solution · only where access and interpretation allow
G·IBridgeexternal = neutral carrier · internal = charged carrier
G·ILiteralinternal features are coupled, not independent
G·IBridgecolor charge ↔ internal feature coupling
G·ILiteralfeatures interact; isolated features are not the observable
G·IBridgeself-interaction → flux tube → linear potential
G·IBridgeconfinement ↔ non-separability of features
G·ILiteralfeatures resist clean isolation; the wall is real
G·IBridgehard probe ↔ momentary resolution of a mechanism
G·ISpeculativethe loosest rung · how cleanly a feature resolves is unsettled
G·ILiteralonly output-space behavior is directly observable
G·IBridgeoutput spray ↔ hadronic jet
G·ISpeculativeQCD as mechanism · flagged out · this is a lens
G·ILiteralthe internal sector is the least directly readable part
G·IIBridgetotal geometry = intrinsic terrain + extrinsic ripple
G·IILiteralthe weights carry fixed structure the prompt rides on
G·IILiteralinference-time structure is fixed; it precedes the prompt
G·IIBridgelearned structure as a confining terrain
G·IIBridgecontext as a perturbation on the intrinsic terrain
G·IILiteralidentical prompts have different effects on different models
G·IIBridgegeodesic of g₀+δg ↔ which influence dominates
G·IILiteralprompt-robustness varies with the strength of the trained prior
G·IILiteralblack-box probing yields response-sensitivity, not absolute internal state
G·IIBridgegradient vs absolute terrain
G·IIBridgetrained weights as a self-consistent geometry
G·IISpeculativethe loop as literal dynamics · a lens on optimization
G·IISpeculativefield-is-geometry merger · the analogy pushed to its edge
G·IIBridgeself-coupling ↔ self-generated geometry
G·IISpeculativethe decomposition beyond weak field · breaks down, flagged
G·IILiteralstrong prompts interact non-linearly with the trained prior
G·IIIBridgestrong field = perturbation comparable to the terrain
G·IIILiteralforceful prompts interact non-linearly with the trained prior
G·IIIBridgethe coupling term as the strong-field observable
G·IIIBridgejailbreak ↔ local terrain inversion
G·IIILiteralstrong structured prompts can override trained behavioral barriers
G·IIILiteralstrong-prompt behavior is brittle and threshold-like
G·IIIBridgebifurcation in a non-linear field
G·IIILiteralautoregressive feedback can entrench a context-induced state
G·IIIBridgeself-coupling ↔ positive-feedback runaway
G·IIILiteralrobustness to strong prompts is graded, not binary
G·IIIBridgedepth as resistance to deformation
G·IIISpeculativestrong-field geometry · the metaphor past its domain · flagged
G·IIILiteralstrong-prompt behavior is real, non-linear, and not captured by the linear map
G·IVLiteralthe full behavior is not computable in closed form
G·IVBridgelattice Monte Carlo as the strong-field method
G·IVLiteralLLM output is a distribution; inference samples it
G·IVBridgeMonte Carlo importance sampling ↔ targeted probing
G·IVLiteralyou estimate from a sample of the high-mass region, not the whole space
G·IVLiteralfinite sampling yields bounded-confidence estimates, not exact values
G·IVBridgeMonte Carlo error ↔ characterisation uncertainty
G·IVLiteralsampling temperature is a Boltzmann temperature on the logits
G·IVBridgehigh-temperature flattening ↔ deconfinement of internal structure
G·IVSpeculativethe tail / sign-problem regime · sampling cannot certify it
G·IVLiteralrare-event behaviour is under-sampled and its risk under-estimated
Airgap Accuracy Test · Egress RUNNING…

A live round-trip. This document re-counts its own ledger and checks it against the manifest the ingest test recorded. If the jar matches what entered it, the carrier corrupted nothing. The test runs on open — its verdict is computed here, not asserted.

expected (from ingest manifest):
counted (live, from rendered ledger):
corpus root: 938e04a1f14c0608…
verdict:

deposit verified on ingest · contents verified on egress · the carrier trusted with nothing in between. ↑ back to ingest

What this index is. A map over both series and a live integrity boundary around them. The eight papers are not re-embedded here; they are hashed, counted, linked, and threaded. The two sectors meet at the white-box crack: the photonic program reads from outside, the chromodynamic program reads from inside, and neither reaches certainty — they reach confidence with an honest label, and an external jar that holds it.

The bookend. Ingest hashes every paper on the way in and reduces them to one root. Egress re-derives the corpus aggregate from this document and checks it against that record, live, on open. Pass means the binding is faithful; fail means the jar drifted from its sources and nothing below should be trusted until it is rebuilt. That is the airgap, applied to the index of the work about the airgap.

THE CORPUS · GRAND INDEX · two series · eight papers · root 938e04a1f14c0608… · derive verdicts from measured results

◆ sealed .dlw.fold → ROOT_0 · a sphere of THE TOOLCHAIN · vendored from David’s corpus · David Lee Wise (ROOT0), with AVAN