Claims ledger

Any adaptive system that must compress multiple inputs into an irreversible decision under finite resources converges on the committed-node architecture.

Evidence

Stated as the theorem driving the work. Not proven. Fit across domains is an argument, not a test: the series' own rule is that a worked example that fits is the cheapest thing in science to produce.

What would defeat it

A system meeting the antecedent whose decision step is not rate-limiting and irreversible. Or the cross-domain test the series says it has never run: predict a node's type in one domain from determinants that fixed it in another, then measure, and miss.

Source

Theorems page; style sheet v22 §1.2 Definition

A sensor node failing loop closure or steepness queues instead of folding. Relaxation diverges superlinearly (exponent in (1, ∞)) against a fold's [1/2, 1).

Evidence

Derivation in Substrate §8.7. Exponent test is kill test 6 in Dropzone §9.

What would defeat it

A node meeting all three bifurcation conditions whose exponent falls outside [1/2, 1). Or a node failing (i) or (ii) that shows hysteresis.

Source

Style sheet §1.2; Substrate §8.7; Dropzone §9 Definition

In a mean-field (equal-load-sharing) fibre bundle the relaxation time diverges toward failure with exponent 1/2 (analytic); −0.42 measured.

Evidence

DiscriminatorStress §3, N = 10⁶ fibres. One model family, simulated.

What would defeat it

A measured exponent that stays away from 1/2 with the fit window held entirely above the sweep-rate floor.

Source

CNBF-DiscriminatorStress-v1 §3 Definition

In a local-load-sharing bundle (stiff to rupture) relaxation is flat down to 10⁻⁵ of the failure stress. 86% of fibres are intact at the snap and the largest precursor burst in about 11,000 was 10 fibres.

Evidence

DiscriminatorStress §3A, N = 4096. One model family, simulated.

What would defeat it

A hinge-typed model with a scale-free precursor distribution, measured above the event-count floor.

Source

CNBF-DiscriminatorStress-v1 §3A Definition

A hinge announces itself with a scale-free cascade and falling recovery time; a fold with Gaussian fluctuation of rising variance.

Evidence

Withdrawn 2026-11-06. Kill test 1 fired on the pressure test. The kurtosis discriminator failed (2 of 28 cells).

What would defeat it

Withdrawn, not reversed. Restored for a class only if a hinge-typed model shows scale-free precursors above the floor. Typing now falls to the structural channel.

Source

Erratum 2026-11-06 Definition

A cascade is the event-stream view of a fold in a coupled population at gain G approaching 1. Mean burst size 1/(1−G); sizes follow S^(−3/2) at G = 1.

Evidence

Gain read off the event stream matches the formula within 2.2% across seven loads in the mean-field bundle.

What would defeat it

Gain departing from the linearised value beyond finite-N scatter while bursts stay scale-free; or a population at gain near 1 with no scale-free bursts.

Source

Erratum 2026-11-06 Definition

Sweeping the knob too fast mistypes a fold as a hinge: the apparent exponent is biased low (0.117 to 0.393 against a true 0.5). In collapsed units the bias is a calibration known in advance.

Evidence

DiscriminatorStress §3C and §3C.1: 0.4948 over a window of z in [5, 50], zero spread across five decades of sweep rate. Saddle-node only; cusp and transcritical are unchecked.

What would defeat it

An exponent in collapsed units that departs from the predicted value for its window.

Source

Style sheet §1.5; DiscriminatorStress §3C Definition

For a hinge-typed node with no early warning, respond by changing structure (a cap, buffer or rule), not by trying to forecast. "Bound, don't forecast."

Evidence

Argument. Survives the erratum. Addendum status: retrospectively checked, not yet run through a domain kill-test.

What would defeat it

A hinge found to carry a genuine lead-time precursor (retype it dual). Or a structure adopted on this logic that does not bound the loss.

Source

Style sheet §1.1, General Addendum §2 Definition

Recovery does not retrace failure, and the required over-correction grows as the node becomes more switch-like.

Evidence

A prediction from the fold geometry (Substrate §8.5). Apollo 13 on this site is an illustration only.

What would defeat it

Re-challenge leg: restoring the original input restores the original state in a node meeting the three conditions. Withdrawal alone is not a test.

Source

Style sheet §1.5 Definition

Capacity ⇔ hysteresis: a node driven to its limit gives two states if it meets all three bifurcation conditions, and a queue if it does not.

Evidence

Stated in the corpus as asserted and not yet tested. Negative control assigned (DES1 via oxygen). Positive arm unassigned.

What would defeat it

A node meeting all three conditions, driven to capacity with a binding knob, that queues instead.

Source

Style sheet §1.5 Definition

Bistability is erased wherever ΔV < D ln(t_rel/τ₀). The erased set is a collar eroded inward from the whole wedge boundary, not a band at the critical aperture.

Evidence

Derived at style sheet §1.8, with the collar width given there.

What would defeat it

A node below the floor that is observed occupying two distinguishable states.

Source

Style sheet §1.8 Definition

Typing a node rests on the structural channel (configuration, read before events). The kinetic channel corroborates and does not decide.

Evidence

After the erratum, burst statistics are non-identifiable: geometry and coupling combine in the distribution.

What would defeat it

Cases where the kinetic channel alone types nodes correctly, and the structural channel mistypes them.

Source

Erratum 2026-11-06; style sheet §1.5 Definition

Birth is an absorbing transfer edge crossed on purpose, with three properties: absorbing, bounded gap, hand-off.

Evidence

n = 1. Null arms recorded: hatching, germination, the dove (null for property 3b), the tsetse fly. 3b is better stated as withdrawal-gated than completion-gated. The ratchet across births is a candidate only.

What would defeat it

A direct test of whether suckling maintains postpartum uterine tone; further null arms.

Source

CNBF-TransferEdge-v1 Definition

A coupled pair that loses lock before the structure it sits on fails could serve as an endogenous sentinel.

Evidence

A parked angle. The Odd Sympathy story assumes it; it is not shown.

What would defeat it

A real sentinel that fails to lose lock first, or loses it too late to act on.

Source

CNBF-EndogenousMeasurement (working notes) Definition

Prevailing ECS models omit the nutritional substrates that sustain the system. A nutritionally complete model is needed.

Evidence

Preprint 4 (Hypothesis). Hypothesis-generating by its own statement.

What would defeat it

A demonstration that substrate supply has no bearing on ECS signalling capacity.

Source

Preprint 4

Acidic phytocannabinoids are a candidate vitamer family for ECS support. Against eight vitamin criteria: three met, one partly, one undetermined, three not met, and the three essentiality criteria fail together because no deficiency phenotype has been characterised.

Evidence

Preprint 5. The authors do not claim these compounds are vitamins. The paper predicts effect size proportional to baseline deviation from set-point, not to dose.

What would defeat it

Effect size that scales with dose independent of baseline deviation. Or a characterised deficiency phenotype: that would move the essentiality criteria, not close the question.

Source

Preprint 5

Gastric decarboxylation of acidic cannabinoids is negligible. At blood pH they are over 99% carboxylate anions and are conjugated intact (UGT1A9).

Evidence

Preprint 3: a pharmacokinetic argument. What fraction reaches a target tissue is stated as open.

What would defeat it

Measured, significant decarboxylation in the stomach or blood.

Source

Preprint 3

Hemp seed supplies a balanced omega-6 to omega-3 ratio, GLA, complete protein and cofactors; a plausible functional food for ECS support.

Evidence

Preprint 1 says "plausible". The phytosterol mechanism is flagged there as not directly tested.

What would defeat it

Trials showing no change in ECS-relevant measures. None recorded here.

Source

Preprint 1

Hemp seed with raw cannabis leaf and flower aligns with human breast milk across 20 ECS-relevant nutritional axes (Preprint 7, no conclusions drawn). Preprint 8 reads this as an adult formula.

Evidence

The alignment is a comparison of composition. Ordering alone is arithmetic; the stronger form needs a ratio fixed before the comparison.

What would defeat it

A ratio fixed in advance that the composition fails.

Source

Preprints 7 and 8

Much of the apparent contradiction in the THCA-A and CB1 binding literature traces to measurement, purity or citation defects, not to biology.

Evidence

The Perspective: a historiographical reconstruction. Central case: an affinity value attributed to Verhoeckx et al. (2006) from an outsourced assay never published with data or method. The paper does not argue that acidic cannabinoids are CB1 or CB2 ligands.

What would defeat it

A reconstruction showing the case citations did derive from published data.

Source

Perspective

Eating raw cannabis leaf and flower with hemp seed supports the ECS in a way smoking does not ("You don't smoke it. You eat it.").

Evidence

The book's thesis. The author's own Preprint 5 reports that the essentiality criteria are not met. Read this as a hypothesis, not an instruction. Nothing on this site is medical advice.

What would defeat it

The deficiency and dose-response tests Preprint 5 specifies.

Source

Cannabis Is Food (book); Preprint 5

The most upstream determinant of whether an institution takes the adaptive or the collapse branch is the legibility of a defection, scored at ingestion, not the size of the stress.

Evidence

The Legibility paper. A port of the metabolic architecture; the paper asks whether it is a portable scientific object or a coincidence. Type assignments are predictions.

What would defeat it

The paper lists its own refuting tests, falsification first. This site does not restate them.

Source

CNBF-Legibility Definition

Where a node's geometry offers no reliable precursor, the answer is structural: a bound denominated in a variable independent of the one that sets the threshold.

Evidence

Legibility abstract. Same status as L08.

What would defeat it

See L08.

Source

CNBF-Legibility Definition

Authority at one scale is citizenship at the next.

Evidence

Addendum status: retrospectively checked against Substrate and Dropzone, not yet through a domain kill-test.

What would defeat it

A node whose local internal state determines the node above it in a way irreducible to its aggregate output.

Source

Style sheet §1.1, General Addendum §1 Definition

The universe shall have balance.

Evidence

A stance that drives the work. Not tested. Not testable as stated.

What would defeat it

Not applicable. It is labelled so no one mistakes it for a finding.

Source

Theorems page

Life cannot thrive without both substrate and signal exchanged.

Evidence

Proposed label: a principle at n = 1 (the ECS case).

What would defeat it

A life form that thrives without exchanging substrate, or without exchanging signal.

Source

Theorems page

Hemp seed plus raw cannabis leaf and flower is the adult counterpart of human breast milk across twenty endocannabinoid-system-relevant nutrition axes (an "infant formula" and an "adult formula"), including inverted protein and anti-inflammatory entries and a shared dependence on an outside source of vitamin B12.

Evidence

Two observational preprints (Floyd and Benjamin, Preprints 7 and 8) compare published nutrient values side by side and read the pattern as evolutionary. The authors took no measurements. The values have not been rechecked cell by cell for this site, and the author plans to revisit both preprints for rigor. Hemp sprouts have yet to be examined.

What would defeat it

A rerun with every value taken from primary sources for both foods in which the inversions or the shared gap do not hold, or in which the match is shown to depend on which axes were chosen.

Source

Preprints 7 and 8 (doi:10.5281/zenodo.19066408; doi:10.5281/zenodo.20437815); Cannabis Is Food, Chapters 2 and 5 |

Trust is read as a trajectory, meaning its rate and direction of change, and not as a score: two people with identical records can be on opposite paths, and only the motion distinguishes them.

Evidence

The principle stated in the NOP Vision document, section 4, which is the author's own design document for a venture. The author holds a financial interest in that venture (see About). It is a design principle and has not been run through a kill-test. The nearest test the corpus specifies is Kill-test 1 in the Legibility paper: does the same unit raise its own setpoint after a metered stress, read net of membership change.

What would defeat it

A per-unit record in which level alone predicts later failure as well as level plus rate of change does. Or one in which apparent robustness comes only from members leaving (Legibility, Kill-test 1).

Source

NOP Vision section 4; CNBF-Legibility

A referee can promote fair dealing by repricing instead of policing: escrow that releases payment only on verified work, reputation staked against each deal, and verification of claimed skill are meant to make honest work pay better per joule than fraud. Joule is NOP's unit of exchange.

Evidence

A design rule from the NOP origin chat session ("trust thermodynamics"), the author's own venture. The author holds a financial interest in it (see About). Not tested on this site.

What would defeat it

A running market, or a run of the rules, in which fraud still earns more per joule than honest work once escrow, staking and verification are in force.

Source

NOP origin chat session

Below a minimum amount of observed work, a trust engine should report insufficient evidence, not a low score. A newcomer is not a low-trust member.

Evidence

The applicability floor of the Legibility paper (section 6): below a sampling density the instrument is silent, not wrong. Applied to NOP in a July 2026 review of the Trust Algorithm Spec, which had no such floor. The spec has not progressed and is to be rebuilt. The author holds a financial interest in the venture (see About). Not tested on this site.

What would defeat it

A record in which units below the floor later fail at a higher rate than established ones, so that withholding a score loses information a low score would have kept.

Source

CNBF-Legibility section 6; NOP Trust Algorithm Spec review (July 2026) Definition

If acidic phytocannabinoids normalize state-dependent signaling toward an endogenous set-point, their effect should scale with distance from baseline and reverse direction across the set-point, where a full agonist's effect tracks dose.

Evidence

Preprint 5 (version 2), section 3.4. The paper says a gradient in size of effect alone cannot separate this from a plain model in which more tone is worse, because both predict that effect tracks baseline. Only the sign reversal separates them. The paper makes no recommendation and no CB1 or CB2 claim. Not tested.

What would defeat it

A uniform, one-directional response regardless of baseline, in either arm of a placebo-controlled trial that enrolls subjects on both sides of the set-point and analyzes the placebo arm the same way, to exclude regression to the mean (Preprint 5, section 6).

Source

Preprint 5 (version 2), sections 3.4 and 6