Universal Natural Intelligence

This is the document. Rendered from the repository at the commit above, with nothing rewritten for the web. A gate re-renders it on every deploy and fails the build if a single byte differs.

RENDERED FROM STRUCTURED DATA — this page is not a markdown file. Source: lexicon/CONCEPTS.json in uni-cookbook at 575fc93d9d31. The sha256 recorded for this page is the digest of that JSON, not of this text. The renderer is generators/derive_docs.cjs; it copies fields and adds no facts. Every value below is a field of the source, verbatim, including the ones that read NOT RUN and PENDING.

The canonical concept list: the backbone of the five-register lexicon (Sanskrit / Latin / English / Hindi / Spanish). Each entry fixes ONE concept by its DEFINITION, not by its word. The definition is the translatable object and the test object: a candidate rendering in any register is admissible only if it back-translates to a sentence that preserves this definition's discriminating property, and a translator is tested on the definition, never on the headword. NOTHING HERE IS TRANSLATED YET. This file carries no Sanskrit, Latin, Hindi, or Spanish, and asserts no equivalence in any register; per rule 18.14 of the operator's own charter ('Do not claim perfect translation'), every future rendering enters PENDING / SUB EXAMINE and is admitted only against a cited dictionary headword plus a passing back-translation test. Sanskrit technical vocabulary for modern math and physics largely DOES NOT EXIST as attested classical terms; the honest expectation is many coined entries, recorded as coinages, and no attested word may be pretended to mean 'expected free energy'. Sanskrit philosophical vocabulary carries DOCTRINAL FREIGHT (a Samkhya or Vedanta term imports a metaphysics); the ambiguity_notes of the future register files, not this file, are where that is recorded, and no translation choice may quietly assert a metaphysics. The 'store' field routes each concept to its sovereign store and is NEVER a claim about the concept's truth: TRUE = the concept names something falsifiable/reproducible/recalibratable; HONEST = it names lived experience, which is never calibrated; META = it names commentary about signals, which travels separately and is never True. Two counts are load-bearing and are reported here rather than discovered later: GENERATIVE_PROCESS appears in 0 files of this corpus (grep, 2026-07-15) and is introduced here because the model/process conflation is the field's most-confused pair; META appears in 0 files of this corpus and is carried from the wider UNI Tree doctrine (arca META), not from anything the encyclopedia says. This file is a lexicon backbone. It raises no UNI rung, discharges no gate, and states no fact about nature. A nature citation is never a UNI gate.

Field contract — [object Object]

active-inference-math — 29

concept	English	definition
`ACTIVE_INFERENCE`	active inference	A scheme in which perception minimizes variational free energy over beliefs and action minimizes expected free energy over policies, so that perceiving and acting are one operation applied through the two halves of a blanket.
`VARIATIONAL_FREE_ENERGY`	variational free energy	A scalar functional F[q,o] = E_q(s)[ln q(s) - ln p(o,s)], in nats, that scores a belief q(s) against observations o and upper-bounds surprisal, so minimizing it over q both drives q toward the true posterior and tightens the bound.
`EXPECTED_FREE_ENERGY`	expected free energy	A scalar G(pi), in nats, that scores a candidate policy rather than a belief, equal to risk plus ambiguity and equivalently to the negative sum of epistemic and pragmatic value, so that the lowest-G policy is the chosen one.
`SURPRISAL`	surprisal	The self-information -ln p(o) of an observation under a model, measuring how unexpected that observation was, in nats, and being the quantity that variational free energy upper-bounds.
`GENERATIVE_MODEL`	generative model	The joint distribution p(o,s) that an agent HOLDS over observations and hidden states, specifying how states are believed to cause observations and how states are believed to evolve, and which is a choice the agent makes rather than a measurement of the world.
`GENERATIVE_PROCESS`	generative process	The actual dynamics of the world that really produce observations, which exist independently of any agent's beliefs and need not share the generative model's form, variables, or factorization.
`RECOGNITION_DENSITY`	recognition density (approximate posterior q)	The distribution q(s) that is an agent's honest and improvable belief over hidden states, being the object optimized during perception and generally NOT equal to the true posterior.
`TRUE_POSTERIOR`	true posterior	The exact conditional p(s\|o) implied by the generative model, generally intractable because it requires the marginal p(o) = SUM_s p(o,s), and which q approximates without necessarily ever equalling.
`BOUND_GAP`	bound gap	The non-negative slack D_KL[q(s)\|\|p(s\|o)] between variational free energy and surprisal, which is zero exactly when q equals the true posterior and is unobservable without the very posterior that was intractable to begin with.
`ACCURACY`	accuracy	The term E_q(s)[ln p(o\|s)] in F's complexity-minus-accuracy decomposition, measuring how well a belief explains the observations actually received.
`COMPLEXITY`	complexity	The term D_KL[q(s)\|\|p(s)] in F, measuring how far a belief has moved off its prior and penalizing that movement so that minimizing F selects the simplest sufficient explanation rather than the best fit.
`KL_DIVERGENCE`	Kullback-Leibler divergence	A non-negative measure of the difference between two distributions, zero if and only if they are equal (Gibbs' inequality), whose non-negativity is precisely what makes F upper-bound surprisal.
`ENTROPY`	entropy	The expected surprisal H of a distribution, measuring the average uncertainty of outcomes drawn from it, in nats.
`EVIDENCE`	evidence (marginal likelihood)	The marginal p(o) = SUM_s p(o,s) of observations under a model, whose negative logarithm is surprisal and whose logarithm F lower-bounds in negated form.
`NAT`	nat	The unit of information under natural logarithms, equal to log2(e) or about 1.442695 bits, in which F, G, entropy and surprisal are all measured.
`RISK`	risk	The divergence D_KL[q(o\|pi)\|\|p(o\|C)] between a policy's predicted outcomes and its preferred outcomes, forming one of the two terms of expected free energy.
`AMBIGUITY`	ambiguity	The expected entropy of the likelihood, E_q(s\|pi)[H[p(o\|s)]], measuring how uninformative observations will be even if the policy reaches the states it intends.
`EPISTEMIC_VALUE`	epistemic value	The expected information gain of a policy, being the expected divergence between the belief that would be held after observing and the belief held now, which minimizing G maximizes.
`PRAGMATIC_VALUE`	pragmatic value	The expected log-preference E_q(o\|pi)[ln p(o\|C)] of a policy, measuring how far its predicted outcomes move the agent toward its preferred observations.
`POLICY`	policy	A candidate action sequence pi over which expected free energy is scored and among which one is selected.
`POLICY_PRECISION`	policy precision (gamma)	The inverse-temperature parameter gamma > 0, carrying units of inverse nats, that sharpens the softmax over policies, with gamma approaching 0 giving indifference and gamma approaching infinity giving deterministic argmin G.
`PRECISION`	precision	The confidence weighting on a distribution or signal, formally the inverse variance or the inverse-covariance (precision) matrix, which sets how strongly a term is allowed to drive inference.
`PRIOR`	prior	The distribution p(s) held over hidden states before the current observation is taken into account.
`PREFERENCE`	preference (C)	A prior p(o\|C) over PREFERRED observations that encodes a goal as a distribution rather than as a reward signal, and which breaks the tie between revising beliefs and acting on the world.
`LIKELIHOOD`	likelihood	The mapping p(o\|s) from hidden states to observations, realized as the A tensor in a discrete POMDP.
`TRANSITION`	transition	The mapping p(s'\|s,u) from a current hidden state and an action to the next hidden state, realized as the B tensor in a discrete POMDP.
`ELBO`	evidence lower bound (ELBO)	The negative of variational free energy, being a lower bound on the log evidence ln p(o), such that maximizing it is identical to minimizing F.
`FREE_ENERGY_PRINCIPLE`	free energy principle	The proposal that self-organizing systems at non-equilibrium steady state minimize a variational free-energy bound on surprisal, which designates a MODEL STRUCTURE rather than an empirical finding.
`ACTION_AS_INFERENCE`	action as inference	The consequence of the blanket partition that a prediction error can be resolved either by changing internal states to fit the world or by changing active states until the world fits the model, both minimizing the same quantity.

sovereign-vocabulary — 27

concept	English	definition
`TRUE_SIGNAL`	TRUE signal	A signal that is falsifiable, reproducible and recalibratable, with every calibration present and prior recorded, such that if it cannot be falsified or reproduced it is not True.
`HONEST_SIGNAL`	HONEST signal	Lived experience, respected exactly as lived and NEVER calibrated, which is not a failed measurement but a different kind of signal, and whose lack of a falsifier is the point rather than a gap.
`META_SIGNAL`	META signal	Commentary about a signal, which travels separately from the signal itself and is never True.
`SOVEREIGN_ISOLATION`	sovereign isolation	The rule that TRUE and HONEST signals keep separate stores, separate ledgers and separate Merkle roots, and that crossing them is the cardinal sin.
`SIGNUM_SIGNUM_MANET`	the signal remains the signal	The doctrine that the signal and the commentary about it travel separately, so that an attribution is never merged into the value it describes.
`RES_IPSAE_NON_SIMULACRA`	the things themselves, not simulacra	The doctrine that no value, citation, URL, DOI or dictionary entry is ever fabricated: source it, or write NOT-MEASURED, and there is no third state.
`CARDINAL_RULE`	the cardinal rule	A nature citation is never a UNI gate: reading Kleiber's law raises no UNI rung, and no operation takes a NATURA row and a UNI row and produces a stronger row of either kind.
`UNI_FENCE`	the UNI four-value fence	The vocabulary proven / designed / hypothesized / not-yet-built that governs UNI's own build status and is never used of nature.
`PROVEN`	proven	A held, sealed, UNI-signed PASS (Class A or C) with its falsifier still live -- a status of UNI's own artifact passing UNI's own gate, and reserved for the UNI ledger.
`DESIGNED`	designed	A typed spec or signed-in-principle design exists while the build is partial or unverified, which raises nothing and leaves the rung's status unchanged.
`HYPOTHESIZED_UNI`	hypothesized (UNI fence value)	A stated UNI mechanism with no sealed gate yet (Class U), not claimed.
`NOT_YET_BUILT`	not-yet-built	No engine, no run, no gate: an absence in UNI's build, hard-fenced as north-star framing.
`NATURA_CLASS`	NATURA evidence class	The twelve-class, three-group vocabulary (measured / derived / fenced) that governs nature's observed regularities and is never merged with the UNI four.
`OBSERVED_REPLICATED`	OBSERVED-REPLICATED	Measured and independently replicated across labs, datasets or taxa, carrying a real citation, a falsifier and the known range.
`OBSERVED_SINGLE`	OBSERVED-SINGLE	Measured by a method another party could repeat, but with NO independent replication on record, asserting the measurement and asserting nothing whatever about corroboration.
`OBSERVED_CONTESTED`	OBSERVED-CONTESTED	Measured, but with the community genuinely disputing the value, the exponent or the mechanism, so that both positions must be carried and the dispute named.
`MODELED`	MODELED	Derived from a model that fits data, where the model's ASSUMPTIONS are the fence and must be cited with it.
`MODELED_CONTESTED`	MODELED-CONTESTED	Derived from a model AND genuinely disputed by the community, so the row must survive both tests rather than one.
`HYPOTHESIZED_NATURA`	HYPOTHESIZED (NATURA class)	A mechanism about nature that is proposed but not settled, and is not claimed.
`INADMISSIBLE`	INADMISSIBLE	A claim that failed observation, or that is unfalsifiable as stated, carried WITH the receipt of why it failed -- never asserted, never mocked.
`SUPERSEDED`	SUPERSEDED	A value that was carried in the literature and has since been retracted, withdrawn by its own authors, or contradicted by a calibrated re-analysis, kept as the receipt for why it must not be cited.
`NOT_MEASURED`	NOT-MEASURED	The honest empty face: nobody has measured it -- an absence in the literature, written instead of inventing a number.
`NOT_SOURCED`	NOT-SOURCED	The chapter states a value it could not trace to a source: somebody measured it, and this corpus could not find who.
`NOT_CONFIRMED`	NOT-CONFIRMED	The primary source is named at one remove -- a secondary or tertiary work cites it -- but the primary was not read in this pass, so the provenance chain is known and unwalked.
`NOT_LOCATED`	NOT-LOCATED	A named source was searched for and does not appear to exist -- the strongest of the three provenance fences.
`QUAESTIO_APERTA`	open question	A permanent open question -- 'full human' and 'the next evolution beyond human' -- which is never a target, never a milestone, and never a deliverable.
`NONDUM_FALSIFICABILIS`	not yet falsifiable	A REGISTER STATUS and not an evidence class: the mark of an item held in the open-questions register, outside the Tree, which is read and not acted upon.

method — 21

concept	English	definition
`FALSIFIER`	falsifier	A stated observation that would refute a claim, written before the run, without which surprise cannot be computed and the claim is inadmissible.
`RECEIPT`	receipt	The captured observation that closes a turn of the loop -- an audit id and UTC for live state, file:line for code, a run log for tests -- against which the prior prediction is compared.
`EVIDENCE_CLASS`	evidence class	The A-to-U label recording HOW a claim is known, above which no claim may ever be raised.
`CLASS_AUTHORITY_ORDERING`	class authority ordering	M9's rule that Class-B tool state overrides Class-G own narrative and Class-A runtime observation overrides Class-E test pass, so that a passing test does NOT satisfy a criterion demanding a runtime observation.
`LEDGER`	ledger	The append-only single source of truth whose four states are PASS, FAIL, NEGATIVE and PENDING, and which wins wherever prose disagrees with it.
`CALIBRATION`	calibration	Moving wording DOWN to the measured value -- never up -- and keeping the prior wording on record, so that the fence gets louder under pressure rather than wider.
`NEGATIVE`	negative	A held UNI run that missed its pre-registered bar: a result UNI earned, published front-and-center as the credibility, and never retried until it wins.
`BASELINE`	tuned baseline	A strong comparison tuned with the same effort spent on the method itself, since an untuned baseline measures the author's enthusiasm rather than the mechanism.
`DISCRIMINATOR`	load-bearing discriminator	A manipulation -- shuffling labels, swapping markers, ablating the mechanism to zero -- that must COLLAPSE the claimed gain, since a gain surviving the destruction of its own mechanism was never the mechanism's.
`ABLATION`	ablation	Removal of a mechanism in order to measure its contribution, whose number must be a computed residual and never a literal typed in by hand.
`PRE_REGISTRATION`	pre-registration	Writing the bar, the margin, the threshold, the named ablation and the stopping rule BEFORE the build, because reasonable data-contingent analytic choices inflate false positives with no deliberate p-hacking at all.
`HELD_SET`	held set	Data touched exactly ONCE behind a seal-before-scoring step and a once-only sentinel, since a held set scored twice is a training set with a good reputation.
`BOUND_NOT_POINT`	the bound, not the point estimate	M2's verdict rule: the verdict is the confidence-interval bound that EXCLUDES the threshold, never the point estimate.
`SEVERITY`	severity	Mayo's requirement that a passing test counts as evidence only if it could have failed, made operable as the single question: if the claim were false, what is the probability this exact procedure would still have passed it?
`STRONG_INFERENCE`	strong inference	Platt's procedure of devising alternative hypotheses, devising a crucial experiment each of whose possible outcomes EXCLUDES one or more of them, carrying it out to a clean result, and recycling.
`MULTIPLE_WORKING_HYPOTHESES`	multiple working hypotheses	Chamberlin's remedy of plurality rather than neutrality: keep at least two hypotheses always and at least three for anything you intend to claim, each structurally distinct in mechanism rather than in parameter.
`REPLICATION`	replication	Independent repetition of a measurement by another party, whose presence or absence is exactly what separates OBSERVED-REPLICATED from OBSERVED-SINGLE.
`ANOMALY`	anomaly	A residual that survives the hunt for instrument error -- stable, isotropic-or-not, and quantitatively characterized -- which is the most valuable object in science and the seed of the trajectory nuisance to residual to characterized anomaly to discriminating test to precision measurement.
`POSITIVE_CONTROL`	positive control	A run proving the instrument can detect the effect it is looking for, without which a negative result is a hypothesis rather than a finding.
`PROVENANCE`	provenance	The record of what was measured, on what instrument, by whom, when, and through which cable, kept well enough that the loose connector can be found.
`CAVITY_PRINCIPLE`	the cavity principle	M22's rule that a hierarchy level must never treat an upstream prior as fresh evidence, but must divide it out.

natura — 20

concept	English	definition
`NATURE_AS_AUTHORITY`	nature as the authority	The doctrine, defensible in exactly one form, that nature is authoritative because it is the only system that has already run the experiment -- a very long parallel search under real physical constraints in which the failures were deleted.
`ADAPTATIONIST_COUNTERWEIGHT`	the adaptationist counterweight	Gould & Lewontin's mandatory companion to nature-as-authority: not every trait is an adaptation, because phylogenetic inertia, drift, developmental constraint, pleiotropy and historical contingency all produce features that are not optimal solutions to anything.
`CONVERGENT_EVOLUTION`	convergent evolution	Independent lineages arriving at the same solution, which is EVIDENCE of a constraint-optimum and never proof of one.
`CONSTRAINT_OPTIMUM`	constraint-optimum	A solution repeatedly rediscovered by independent lineages under real physical constraints, which is what convergent evolution is evidence of.
`RATIO`	ratio	A quotient of two measured quantities, which usually does NOT survive a change of scale and which must carry its scope, units and class to be admissible.
`DIMENSIONLESS_NUMBER`	dimensionless number	A quotient whose units cancel, which travels across scale precisely because it does, and which is therefore the group to read before any ratio.
`REYNOLDS_NUMBER`	Reynolds number	Re = rhovL/mu = v*L/nu, the ratio of inertial to viscous forces, which spans about thirteen orders of magnitude from bacterium to whale and across which the physics of swimming changes IDENTITY rather than degree.
`SCALING_LAW`	scaling law	A power law fitted within a regime, whose exponent stops being a fact about the world at a regime boundary and which does not extrapolate through one.
`ALLOMETRY`	allometry	The scaling of a biological quantity with body mass, canonically BMR proportional to M^b, whose exponent moves with analysis choices and changes at the major evolutionary transitions.
`REGIME_BOUNDARY`	regime boundary	The point at which the dominant dimensionless group moves by orders of magnitude or the dominant force changes identity, past which a fitted exponent does not carry.
`SCALE_LADDER`	the scale ladder	The sixteen-rung sequence from quark to galaxy ordered by characteristic length, on which Markov blankets appear at the organelle, cell and organism rungs and vanish above and below.
`RUNG`	rung	A level that exists only when all three of these hold: a boundary with real conditional independence forms, a characteristic timescale separates from the level below, and the level has its own dynamics that do not require tracking the level below.
`EMERGENCE`	emergence	A statement about spectral gaps in a dynamical system: timescale separation makes a level's own dynamics writable without tracking the level below, so that broken symmetry at each level generates concepts not present in the level below.
`FREQUENCY`	frequency	A rate of repetition in hertz, admissible only with its measurand and scope, since values landing in the same numeric neighbourhood may be entirely different physics.
`RESONANCE`	resonance	The selective amplification of a driven system near a characteristic frequency of a real cavity or oscillator, which licenses the frequency and nothing attached to the frequency.
`DIFFUSION`	diffusion	Transport by random thermal motion, whose characteristic time scales as the square of distance and which therefore fails as a delivery mechanism above a characteristic length.
`GRADIENT`	gradient	A spatial variation of a quantity, which is a measured object at one rung and a rhetorical one across rungs unless the measurand, its units, and two rungs measured with it are all stated.
`MORPHOGENESIS`	morphogenesis	The generation of form from local rules during growth, in which pattern arises without a blueprint.
`CELL`	cell	The design unit of biology and the canonical Markov blanket, plasma-membrane bounded and dominated by viscosity and diffusion at Re of order 1e-5.
`MEMBRANE`	membrane	A selectively permeable lipid boundary, which is a surface for reasons of lipid physics and history, and whose satisfying I(mu;eta\|b) = 0 is a DIFFERENT question whose answer is NOT-MEASURED.

markov-blanket — 14

concept	English	definition
`MARKOV_BLANKET`	Markov blanket	The set b = {s,a} of sensory and active states whose conditioning renders internal and external states independent, so that p(mu,eta\|b) = p(mu\|b)p(eta\|b), equivalently I(mu;eta\|b) = 0.
`MIND`	mind (internal states, mu)	The internal states mu inside a blanket, which have no term in external states and whose only evidence about the world is the sensory states.
`BODY`	body (the blanket, b)	The blanket b = {s,a} itself: the surface that carries every dependency between mind and world, and that thereby determines what is inferable at all.
`MIND_BODY`	mind.body (the interface)	The coupling mu to b -- a pair of one-way doors, sensory states the world's only route in and active states the mind's only route out -- which is not a fifth thing but the interface itself.
`WORLD`	world (external states, eta)	The external states eta outside the blanket, which a mind reaches exactly never and which enter only as a hypothesis that explains the blanket.
`SENSORY_STATES`	sensory states	The blanket states s influenced by external but not internal states, forming the world's only route into the mind.
`ACTIVE_STATES`	active states	The blanket states a influenced by internal but not external states, forming the mind's only route out into the world.
`CONDITIONAL_INDEPENDENCE`	conditional independence	The relation mu perp eta given b, which is the whole content of a blanket: given the blanket, internal and external states carry no further information about each other.
`CONDITIONAL_MUTUAL_INFORMATION`	conditional mutual information	The quantity I(mu;eta\|b), in nats, which is the operable test of whether a candidate boundary is a real blanket, since I greater than zero beyond estimator noise means a route from world to mind bypasses the body and the partition is drawn in the wrong place.
`PEARL_BLANKET`	Pearl blanket	A blanket as Pearl coined it: a formal statistical construct over a graph you drew, doing conditional-independence bookkeeping relative to a model, and thereby cheap, well-defined and unobjectionable.
`FRISTON_BLANKET`	Friston blanket	A blanket read as the metaphysical boundary of a real thing -- the claim that some system out there genuinely has an inside, a surface and an outside -- which requires premises beyond the formalism's mathematical success.
`BLANKET_SLIDE`	the Pearl-to-Friston slide	The field's characteristic error of defining a Pearl blanket, which is free, and then quietly concluding one has discovered where the organism ends, which is expensive and requires a measurement nobody made.
`INTEROCEPTION`	interoception	Sensory channels whose external states are other parts of the same organism, forming a blanket nested inside a blanket, and rendered in M12's engineering clause as hardware signals.
`BLANKET_NESTING`	blanket nesting	The recursion in which one level's blanket becomes part of the next level's internal states, running organelle to cell to tissue to organ to organism to colony to ecosystem.

pomdp-tensors — 9

concept	English	definition
`POMDP`	partially observable Markov decision process	A discrete formalism of hidden states, observations and actions in which states are not directly visible, over which this corpus's engine runs the loop perceive, EFE-plan, act, learn.
`A_TENSOR`	A tensor (likelihood)	The tensor encoding p(o\|s), mapping hidden states to observations, and one of the four tensors this corpus's engine learns by conjugate-Dirichlet counts.
`B_TENSOR`	B tensor (transition)	The tensor encoding p(s'\|s,u), mapping a hidden state and an action to the next hidden state, and one of the four tensors learned by conjugate-Dirichlet counts.
`C_TENSOR`	C tensor (preferences)	The tensor encoding the prior p(o\|C) over preferred observations, which -- alone among A, B, C, D and E -- is SET rather than learned by Dirichlet counts, being fixed in this corpus by the Z affect modulator.
`D_TENSOR`	D tensor (initial state prior)	The tensor encoding the prior over hidden states at the start of a trial or episode, and one of the four tensors learned by conjugate-Dirichlet counts.
`E_TENSOR`	E tensor (habits / policy prior)	The tensor encoding the prior over policies, biasing policy selection before any expected-free-energy score is computed, and one of the four tensors learned by conjugate-Dirichlet counts.
`DIRICHLET_LEARNING`	conjugate Dirichlet learning	The exact update of the A, B, D and E tensors by counts + lr * sufficient_stat, which is conjugate and therefore requires no gradient, no autodiff, and no backpropagation.
`HIDDEN_STATE`	hidden state	The quantity s that an agent cannot observe directly and must infer, and about which the recognition density is a belief.
`OBSERVATION`	observation	The quantity o that actually arrived at the sensor, which is what a belief is scored against and what a receipt records.

build-objects — 7

concept	English	definition
`AGENT`	agent	A system that holds a generative model and minimizes free energy through a blanket, concerning which nothing further is implied: a system that minimizes F is a system that minimizes F.
`ORGAN`	organ	A named instrument built to run one loop, carrying its own harness, and durable only once committed rather than once deployed.
`ORGANISM`	organism	The scale-ladder rung at which a Markov blanket is identifiable, gravity enters as a dominant constraint, and metabolic rate scales with body mass within its regime.
`HIERARCHY`	hierarchy	Nested levels each of whose blanket becomes part of the next level's internal states, made reachable by evolution because stable intermediates are what interruption spares.
`WORLD_COUPLING`	world coupling	The interface by which an engine's internal states are exposed to a world, isolated by explicit whitelist rather than by blacklist.
`Z_MODULATOR`	the Z affect modulator	The global vector of energy, arousal, valence, fatigue, pain, threat, safety and inflammation that sets precision, preferences, habits, learning rate and horizon.
`SENSORIUM`	sensorium	A body-to-mind categorical observation interface whose modality count is an empirical question, kept at seven because an over-compressed two-modality bottleneck went NEGATIVE on held data.

world-context — 3

concept	English	definition
`GAIA`	Gaia	The claim that the biosphere homeostatically regulates the planet, which is HYPOTHESIZED as regulation with a live unanswered unit-of-selection objection, and INADMISSIBLE as the literal claim that the Earth is an organism.
`LABORATORY`	laboratory	The bounded setting in which a pre-registered metric and a tuned baseline decide a design, and whose engineering or substrate evidence is never general evidence about the science.
`TOY_WORLD`	toy world	The honest scope of this whole program: a developmental active-inference SIMULATION, a bounded peek, a toy world, and never a person.

sha256 4bcf4fa538ef68a4 — of the original file, so what was ingested stays checkable.

The lexicon — every concept, by domain

How to read this page

Precise — the source document