Topics — What a work is about. Nucleation holds whatever reading the owner files here, whether or not an Anthology of the SOTA topic could also hold it; instruction for ML practice belongs to the anthology (ADR-048). The words are seeded from the clusters the reading-time triage of 2026-09-25 found, and grow by decision.
Cognition — the mind as information processing — perception, memory, reasoning and cognitive strategies — as studied by cognitive science and philosophy of mind (group: philosophy).
29 of 547 CLAIM documents. Back to the full index.
| # | Title | Summary | Status |
|---|---|---|---|
| CLAIM-003 | A global workspace need not give every subsystem the same access: access is compartmented in brains as in organizations, so “global” overstates it | The owner’s point at U26: an organization’s permission structure plays the role functional connectivity plays in a brain, and the daydreaming driver shows access within one agent can be compartmented. | Deferred |
| CLAIM-066 v2 | Kirby, Cornish and Smith show structure emerging under transmission with an experimenter’s filter, not filtering by measured inductive biases, and Griffiths and Kalish’s convergence to the prior holds only for sampling learners | From the readings of Griffiths and Kalish (NOTE-599) and Kirby, Cornish and Smith (NOTE-596, Skimmed). The manuscript cites them as showing one thing; they show different things. | Superseded — by ARG-035 |
| CLAIM-068 v3 | Repeated reconstruction is attraction toward conventional regions of communicative possibility: pragmatic stances, not only lexicon and grammar, can be stabilized by transmission | A50 §3 and §6, A52 §5.5 and §7.4–7.5, recovered. The manuscript keeps the stable conventional endpoint and one stance-drift example, but not attraction, selection pressures or the extension of iterated learning to footing. Apparently dropped by inadvertence. | Proposed |
| CLAIM-090 v3 | Serial transmission is reconstructive and filtered by the transmitters’ inductive biases, so a stable endpoint can be far from the ancestral utterance while remaining interpretable | The manuscript’s §8, citing Bartlett, Griffiths and Kalish, and Kirby, Cornish and Smith together as showing the same thing. CLAIM-066 separates what each shows. | Proposed |
| CLAIM-275 v2 | An intervention can come before the interpretation it makes available, so in an organization of processes meaning, policy and influence can co-evolve, and none of them is upstream of the others in every case | A4 §6, on the record’s two accounts of emotion, Frijda’s appraisal upstream of action readiness and Barrett’s construction, and the owner’s two replies at U5: “yes, exactly” to “meaning, policy, and influence can co-evolve”, and “so much yes” to “An operator may act not because it already possesses an interpretation, but because its intervention creates the conditions under which a particular interpretation becomes available.” The first filing left both out. It does not say interpretation never comes first. | Active |
| CLAIM-296 v2 | What a system has, what it knows it has, and what it can deploy within a task’s deadline are different, and the last is a relation among the resource, the agent, its environment, the task, the deadline and the cue, not a property of any one of them | A18’s first candidate, from the owner’s EMT case (CASE-085): the drug was stocked, the EMT knew it, and it could not be found before the medic moved on. “If I can’t find it, it isn’t there” is false of the inventory and right of the decision. TERM-055 names the relation. A19 added the cue: the same memory under a different request may retrieve it. | Active |
| CLAIM-298 v2 | A frame can change a judgement by changing which of an interpreter’s stored distinctions are reachable, without supplying any evidence about what is being judged | A20’s proposed claim, after the owner tied addressing to anchoring at U20, and A19’s “addressing conditions”. Its psychological form is the selective-accessibility account of anchoring, and the readings of 2026-10-15 weaken it: Strack and Mussweiler (1997, read) infer accessibility from moderations and latencies without measuring it, and their model lets the same retrieved knowledge push a judgement either way; Harris et al. (2019, read) find the most direct measure of the accessibility step does not replicate and cannot detect priming at all. The retrieval step is inferred, not shown. The mechanism the record holds, attention as sparse distributed memory (LIT-269), is a model, not evidence about judges. The extension to communication in general is the record’s and is not shown. | Proposed |
| CLAIM-306 | Under a deadline, search can establish presence by finding one item but absence only by exhausting the collection, so “If I can’t find it, it isn’t there” acts on an unfinished search as negation as failure acts on a finished one: a default stronger than either the closed-world assumption or negation as failure, safe only when the arrangement makes the full search finish within the deadline | The medic’s “If I can’t find it, it isn’t there” (CASE-085) and the owner’s “if I can’t find it quickly enough, it may as well not be there” (CASE-094) act on a failed search as if it showed absence. One find shows presence, but absence needs a check of everything, which a deadline forbids. First filed as negation as failure on a time bound, with Reiter and Clark unread. Both readings corrected it. The two assumptions are separate. Reiter’s closed-world assumption is that the database is complete about the world. Clark’s negation as failure is licensed only by finite failure: “the exhaustive but unsuccessful search for a proof”. A search cut off by a deadline meets neither, so the rule is stronger than both. It assumes a closed world and that the unsearched part would have failed. It is safe when the full pass finishes in time. In CASE-085 the EMT knew the drug was stocked, which is exactly where Reiter shows the closed-world assumption inconsistent. So there it was a decision to stop, not an inference. | Superseded — by CLAIM-312 |
| CLAIM-307 | Retrieval pressure and evidence are different criteria for carving a category, and they move in opposite directions with copy count: evidence that a distinction exists grows with its copies, the search a bin saves shrinks with them, so the bins a deadline forces are often the ones the data least supports | From MOM-CLM-166, which reads the owner’s LEGO parable (CASE-094) against a formalism that cited it with the sign reversed. CLAIM-154 says when a distinction is real enough to name. Its evidence grows with the number of copies. The parable says when a distinction must be built to act in time: a single brick among similar ones. Expected search falls as copies grow. The two answer different questions, “is it there?” and “can it be reached in time?”, and come apart in both directions. MOM’s table was re-run here and reproduced, and a simulation of the search agrees. | Active |
| CLAIM-309 | How good a mixture is and how discriminable it is from its diluent come apart: under linear mixing a smoothie’s distance from milk only shrinks as milk is added, so a preference that rises and then falls with milk is over the ratio and cannot be a d′, and the signal-detection reading of the owner’s smoothie line reverses one of its directions | The smoothie note in the owner’s LEGO monograph reads the owner’s line (CASE-096) as signal detection. Milk is the noise, “Less Good” is d′ falling as the mixture approaches milk, and “More Good” is d′ rising as a little milk is added. Under linear mixing the mixture’s taste is (1 − f)s + f·m, and its distance from milk is (1 − f)|s − m|. That falls monotonically from the first drop, so discriminability from milk cannot rise and then fall. The note’s own first model is right: goodness is a single-peaked function of the ratio, which is a preference, not a discriminability. MOM-CLM-166 found the same kind of reversal in the formal reading of the LEGO parable. MOM’s record never mentions the smoothie note. | Active |
| CLAIM-310 v2 | Sorting schemes over objects with determinate properties never conflict as descriptions, since the product of any sortings refines them all; they conflict as arrangements, since a fixed number of bins realizes one partition when each item is stored once, and which costs least depends on what is searched for and on what stands out within a bin | What is directionally right in the owner’s LEGO monograph, as Mathematics of Meaning graded it. Its contextuality claim is empty: any sortings of bricks share a joint refinement (MOM-CLM-137), and the events they generate stay Boolean (MOM-ARG-047). Its intuition that builders cannot share a sorting is true of arrangements. One set of bins realizes one partition, the product may need more bins than there are, and the cheapest choice depends on what each builder searches for and on what stands out within a bin (CASE-095). The disagreement is a trade, not a failure of global truth. Sorting on one attribute also turns a search for a combination of features into a search for one feature within each bin, which Treisman and Gelade (LIT-1082, read) show is the difference between a serial scan and one found at a glance. | Active |
| CLAIM-312 v2 | ”If I can’t find it, it isn’t there” asserts operational unavailability, that the item will not be put to use before the deadline, not absence: a search that fails by the deadline establishes unavailability outright, and licenses absence only with a certificate that the unsearched remainder cannot hold the item, from exhaustion, from an invariant of the arrangement, or from a closed-world default | Restates CLAIM-306 at the owner’s request, after CLAIM-311 showed that absence can be established without exhausting a collection. The owner’s own wording in CASE-094, “it may as well not be there”, is hyperbole for operational availability (TERM-055), not a claim about inventory. Read so, the rule reports an outcome and needs no default. CLAIM-306’s analysis survives as the account of the literal reading: what a failed search must be joined with before it licenses absence. The gap between the two readings is CLAIM-306’s Bayes term, and it matters when the statement is passed to someone who will act on presence. | Active |
| CLAIM-315 | In comparative anchoring tasks, considering an anchor shifts the later numerical estimate without adding evidence about the target, and the shift is carried by anchor-consistent knowledge made more accessible; the shift is established, the accessibility step is inferred and has not been measured | The empirical half of CLAIM-298, split off under ADR-043’s pilot, which found that claim compound: an empirical claim about judgement tasks and an interpretive extension to communication in general. This half is scoped to the comparative-question paradigm. Its first part is established (THEORY-337); its second is the selective-accessibility account (THEORY-335), whose direct test failed to replicate with a measure that also missed direct priming. CLAIM-298 now rests on it. | Proposed |
| CLAIM-325 | The behavioural data the record has read show no contextuality proper: none of the five kinds of data Dzhafarov, Zhang and Kujala re-analyse is contextual, Wang et al. test none of their 90 language systems for it, and no documented test of pragmatic or meaning judgements finds it at adequate power | Split from CLAIM-219 under ADR-043: the empirical premise of that objection’s dilemma, which CLAIM-201 conceded. Dzhafarov, Zhang and Kujala (LIT-264, read in NOTE-235) find none of five kinds of behavioural data contextual, two of them previously published as contextual; Wang, Sadrzadeh, Abramsky and Cervantes (LIT-851) report signalling in 69 of 90 language systems and test none for contextuality; CASE-053 collects the documented tests on elicited judgements and finds no contextual pragmatic data. It is scoped: CASE-053 records a reliable positive result on instructed matching, so “none in behavioural data” is too broad, and the absence is LIT-264’s working hypothesis, not a proof that such data cannot exist. | Active |
| CLAIM-336 v6 | The pragmatic-transport and will-organization lines describe one model, an organization of operators that act on a shared interpretive state under a context while allocating one another’s influence: a frame is the context under which operators act, a judgement is the response they yield, an action-generating regime is an operator or coalition, and a will-organization is the dynamics that allocates influence among them | The owner’s, U2 of the exchange of 2026-10-16: the pragmatic-transport model and the will-organization model are “both manifestations of the same underlying model”. The mapping is the assistant’s, A6, filed at the owner’s request (“address the still open stuff”). At U8 the owner endorsed it as filed (“Yes, as filed”). It reads both lines into CLAIM-297’s operator model. The sharp edge: it says the lines share one model, not that either line’s results transfer to the other unchanged. Each mapping is a hypothesis until a result on one line is shown to hold, through it, on the other. | Proposed |
| CLAIM-337 v2 | Active inference already runs an internal loop: mental action selects among higher-level cognitive states and sets the precision of lower-level observations, structure learning reshapes the generative model without action, and a judgement is behaviour, so a judgement–observation loop may be action and perception one Markov blanket inward | The assistant’s, A1 of the exchange of 2026-10-16, on active inference’s behalf, strengthened at A4. Sandved-Smith et al. cast mental action as policy selection over higher-level cognitive states; Brouillet and Friston make felt fluency an inference about the precision of one’s own processing; Smith et al. let a generative model grow its own structure by inference and learning alone. With Chalmers’s point that judgements are behaviour, the owner’s judgement is active inference’s action relocated. What it does not reach: the diagram. A redescription with no action on the world concedes that action on the world is not part of the loop. | Proposed |
| CLAIM-338 v2 | The selectivity active inference gives action, sampling what one expects to experience, is played in the coupled meaning–authority model by the authority space: influence decides which operators retrieve and so what the next judgement observes, which makes Friston’s self-confirming sampling the model’s authority-to-interpretation arrow and his agent a coupled loop run without independent scoring | The owner’s, U10 of the exchange of 2026-10-16, on Friston et al.’s “sampling selectively what one expects to experience”: “I think this selectivity maps to the authority space in my coupled model.” Worked out by A10. Selective sampling is the β arrow of CASE-099, authority shaping interpretation through what it retrieves. Policies scored on current beliefs are the γ arrow. So Friston’s agent runs both arrows, and his “this weakness becomes the strength” accepts capture, with the anchor moved outside the loop to selection on whole agents. The sharp edge: the mapping relocates sampling from the world to retrieval, which is CLAIM-337’s move reversed, and it holds only if the agent’s own store can surprise it as the world can. | Proposed |
| CLAIM-339 v3 | A judgement–observation loop cut off from action loses one route for evidence it did not select, and the free-energy principle’s reason for minimizing surprise, that a persisting system must keep its states in bounds, needs action on the world, though a loop can confirm itself with action available too | The assistant’s, A1 of the exchange of 2026-10-16, narrowed at A4. Friston et al. (2010) say active sampling is “mandated by ergodic arguments based on the very existence of adaptive agents”, and the enactive account puts a system’s own norms in its capacity to act against tendencies toward its viability boundary. Kübler and Birbaumer’s extinction hypothesis is the empirical form. As first filed it said a loop without action has a self-confirming minimum. NOTE-766 shows the minimum arises with action available, so the counter now claims only that losing action removes one source of evidence the loop did not select. CLAIM-342 answers it. | Active |
| CLAIM-340 | Scoring operators on predictions committed before the evidence is not by itself Regime B: in the linearized operator model the coupling of interpretation to endorsement has the sign of κ′ − κ̄, so commitment removes capture only when the interpretation bends the reading of the evidence no more than it bends the predictions, and it keeps the learning that revising after the evidence would lose | The formal question QUESTION-121 asked, answered in CASE-099’s linearization (CASE-106): is committing predictions before the evidence equivalent to Regime B? No. The coupling γ is 4ηs(κ′ − κ̄), where κ̄ is how far the shared meaning bends the committed predictions and κ′ how far it bends the reading of the evidence. It is corrective, null or capturing as κ′ is below, at or above κ̄. Strong correction can fail by oscillation when |βγ| ≥ 1 − αδ. Revising predictions after the evidence sets γ = 0 but also stops authority from tracking who predicts well. Linear, at the symmetric state, for squared-error scoring. | Active |
| CLAIM-341 v3 | The loop that keeps a lucid agent’s model in touch with what it models runs between judgement and observation: each judgement reshapes the observer whose later observations it then judges, and action on the world is not part of that loop, so active inference’s action–perception cycle misplaces where the agent is active | The owner’s, U1 of the exchange of 2026-10-16: Friston’s free-energy approach is “generally correct, but the cybernetic model is wrong”; action can be cut out of the loop, which decomposes into “Judgement (qua measurement that generates interpretations) <—> Observation”. At U2 the owner held that the pragmatic-transport and will-organization models are “both manifestations of the same underlying model”. The claim is stated in that model’s terms, as CLAIM-297’s circuit, and in the diachronic form A4 proposed and the owner asked to be filed at U5 (“make the revisions”): a judgement need not produce the observation it judges, but it reshapes the observer that produces the next one. The sharp edge: the synchronic form, in which each judgement produces its own outcome, is not claimed. That is the open route of TERM-039 and CLAIM-127, and the record demoted it once before (CLAIM-109). | Active |
| CLAIM-342 v3 | What keeps a judgement–observation loop from confirming itself is evidence whose arrival or scoring does not pass through the interpretation it tests; action is one source of such evidence, not a necessary one, and a loop that selects its own evidence can lock in with action available | The assistant’s synthesis, A3 and A4 of the exchange of 2026-10-16, filed at the owner’s request at U5. Three results the record already holds. CLAIM-330: capture in the operator model needs both arrows of CLAIM-297’s circuit, and scoring operators independently of the dominant interpretation removes it. CLAIM-141: a loop always has a distributional fixed point, so what binds it is a constraint that prescribes marginals. THEORY-285 (NOTE-766): an active-inference agent locks in with action available, because its own sampling policy withholds the evidence that would correct it. The claim does not say action is unimportant, only that it is not what does the anchoring. | Active |
| CLAIM-358 v3 | Learned local accommodation can reduce the amortized cost of recurrent arbitration: where a recurring conflict among an agent’s processes admits a stable accommodation, settling it among the processes can cost less over repeated encounters than escalating each instance to a higher-level process that mediates or overrides, though local negotiation and the relations it needs are not free | The owner’s, U45 of the exchange of 2026-10-17: “processes which can navigate disagreement amongst themselves require fewer resources from the system than processes which need to activate additional subsystems to resolve disagreement.” A45 put it in conditional form: where a recurring conflict admits a stable local accommodation, learning it amortizes the cost of higher-order mediation. It refines CLAIM-261’s “pays for its unity in effort”: the divided agent pays for escalation. The title stated the owner’s thesis unconditionally until the owner’s review of the filing put it in A45’s conditional form: local negotiation, and keeping up the relations among specialists it needs, have costs of their own. The sharp edge: it does not say local settlement is always cheaper, that an integrated agent works less hard, or anything about authenticity. | Proposed |
| CLAIM-411 v5 | Under the broad, nonphenomenal definition of interpretation the constitutive thesis uses, substantial processing that is not independently shown to be experiential meets the definition, blindsight’s apparently unconscious visual discrimination among it, so interpretation alone does not establish experience, and the thesis must say which further organizational properties, if any, make an interpretation experiential | A3’s “most serious challenge” to the owner’s thesis, which every later reply accepted. Once interpretation is defined broadly and without experience, as the thesis requires, it admits much processing that is not shown to be experiential: “Interpretation ⇏ Conscious experience. At least, not under an unrestricted definition of interpretation.” The objection is that the thesis is incomplete, not that it is false. A7 kept it as the theory’s “open sufficiency condition”. The record already separates access from experience (QUESTION-048, THEORY-577), and the candidate conditions the exchange named (global availability, integration, higher-order or recursive modelling, valuation, perspective) are the question QUESTION-254 asks. | Active |
| CLAIM-412 v9 | Experience is constituted by interpretive organization, not produced by it: an appropriately organized process in which signals acquire significance for a system is the experience, so whether a system can have experience becomes whether it can sustain that organization | The owner’s, U1 and U2 of the exchange of 2026-10-21. At U1 the owner argued that if sensory apparatus is necessary for grounded experience, ungrounded experience is possible and “experience is fundamentally an interpretive act”, and proposed recasting the possibility of experience as the possibility of interpretation. A1 asked whether experience is constituted by interpretation or produced by it; at U2 the owner chose “former reading” and named the brain in a vat, the rubber hand illusion and CBT. A1–A7 refined the thesis without changing it. Interpretation is defined without experience, and it is organization, not judgement. The step from organization to experience is an identity to be defended, not a theorem. Its open edge is the sufficiency condition, which CLAIM-411 presses and QUESTION-254 asks for. | Proposed |
| CLAIM-421 v4 | The cessation of an organizational perspective’s enactment need not imply the loss of the dispositions, capacities or reproductive relationships from which it can be reinstated: an interruption can be of activation (the disposition is held but not engaged or retrievable), of enactment (held, but the material, institutional or social conditions for exercising it are gone) or of reproduction (held by some bearers, but the processes that pass it on are damaged), and temporary unavailability is not extinction, though prolonged interruption of reproduction can end a tradition | From the branch of the exchange of 2026-10-22, at the owner’s U37: the doorway effect as perspective meeting grounded experience, and the perspectival interruptions of the Chinese Cultural Revolution. A37 drew the distinction the record’s account of perspective lacked, between the conditions under which a perspective is acquired or reproduced and the conditions under which it stays available. The edge: a doorway and a political revolution are not the same kind of forgetting, and the claim says only that both separate a capacity’s existence from the availability of the conditions for exercising it. | Active |
| CLAIM-475 v2 | Specifying is itself a resource-bounded realization: formulating a question takes time, concepts and computation as realizing a configuration takes energy and matter, so the questions an agent can ask are a product of what it has already realized; a stabilized specification is past inquiry stored so that later questions become cheaper to ask, and it can make other questions harder | The owner’s, U17 of the exchange of 2026-10-27: “resource constraints impact specification in exactly the same way as they impact realization. I can’t interrogate a question that takes me more time than I have to even ask it.” It extends CLAIM-470: if inquiry is specification, specifying is a process with costs, and the feasible questions are bounded before any answering starts. A17 and A18 drew a consequence the record can use: TERM-065 takes its query family as supplied, and this claim says that family is the output of bounded activity. A stable specification amortizes the work that made it, which is a separation of time scales (CLAIM-478), and entrenches what it makes cheap. | Active |
| CLAIM-514 | World-modeling is a family of resource-bounded capabilities for producing and using predictions about an environment’s possibilities and consequences, each relative to a declared family of questions; a world model in the representational sense is one possible realization of them, competent behaviour can determine environmental structure without identifying any internal mechanism, and producing, maintaining and revising world-modeling capabilities are further achievements, not prerequisites of agency in general | The owner’s preferred formulation at the second review of the filing of the exchange “Analyze nucleation concepts” (2026-10-31). It joins the world-model thesis, which holds that behaviour does not identify a mechanism, and its counter from Richens et al., which holds that behaviour across goals determines the dynamics. Both are kept, with each stated about a different notion of world model. It “integrates the Richens result rather than rejecting it, preserves [the] original critique of folk-model attributions, and connects both to the larger Nucleation program.” Proposed: a definition and an organizing thesis, with the study CASE-187 as its test. | Proposed |
| CLAIM-515 | Whether an organization has a cognitive property is four questions, not one: what information about the property can be recovered from its behaviour, what capabilities it can actually exercise, what mechanisms realize them, and whether it can maintain or transform those mechanisms; the four can come apart, and none should be silently substituted for another | The general principle the owner drew at the second review of the filing of the exchange “Analyze nucleation concepts” (2026-10-31): “When asking whether an organization ‘has’ a cognitive property, distinguish the information recoverable from its behavior, the capabilities it can actually exercise, the mechanisms realizing those capabilities, and its ability to maintain or transform those mechanisms.” It was drawn from the world-model case (CLAIM-514) and judged “sufficiently general—and sufficiently testable—that it deserves an explicit place in Nucleation’s research program”. Proposed and interpretive: its instances are demonstrated case by case, and the generalization is not. | Proposed |
| CLAIM-519 v2 | Competent behaviour does not establish an internal world model: over a fixed interface and a finite horizon, an agent with no model, whose actions are looked up from the history so far, reproduces the interaction distribution of an agent that has one exactly; a policy can identify predictive structure without any particular internal component realizing it, so compression and generalization support comparing candidate architectures, and establishing an internal mechanism needs evidence that discriminates candidate implementations, such as targeted interventions | The owner’s thesis at U20 of the exchange “Analyze nucleation concepts” (2026-10-31): that under the record’s framework a world model is “essentially a ‘folk’ explanatory tool and that no such thing need explicitly be present for the external behavior to be indistinguishable from a comparable black box agent which actually did contain a world model”. A20 agreed and gave the construction, with a limit the claim keeps: the attribution “sometimes … names a genuine, interventionally identifiable computational mechanism”. The construction is Block’s lookup-table argument (THEORY-092) applied to agency, so it is elementary; the thesis is the reading of it. | Proposed |