diff --git a/docs/okn/cognitive-kb.sample.ttl b/docs/okn/cognitive-kb.sample.ttl index 7097648f..d5d223a8 100644 --- a/docs/okn/cognitive-kb.sample.ttl +++ b/docs/okn/cognitive-kb.sample.ttl @@ -94,7 +94,7 @@ kg:dataset dcterms:creator ; dcterms:description "Concept pages on Active Inference and the sciences around it, linked by typed and untyped relations, with citation records and navigation pages."@en ; dcterms:license ; - dcterms:modified "2026-10-12"^^xsd:date ; + dcterms:modified "2026-11-15"^^xsd:date ; dcterms:publisher ; dcterms:source ; dcterms:title "Cognitive Active Inference knowledge base"@en ; diff --git a/docs/okn/semantic-review.json b/docs/okn/semantic-review.json index 306a7675..3d95591d 100644 --- a/docs/okn/semantic-review.json +++ b/docs/okn/semantic-review.json @@ -19691,7 +19691,9 @@ "", "", "", + "", "", + "", "", "", "", @@ -19709,6 +19711,7 @@ "", "", "", + "", "", "", "", @@ -19762,7 +19765,7 @@ "yaml_type": "date" } }, - "source_sha256": "ab80cb6a99b0b89611cce1915c2316ff02b151359261af6afb55a5f29139e90b", + "source_sha256": "64f4b7c5d7fea4e92a00f7677994c91f4d837cce5902c2361ed01874fdf10da5", "source_type_terms": [ "\"index\"" ], diff --git a/docs/okn/semantic-review.md b/docs/okn/semantic-review.md index dfec9700..b010a9d1 100644 --- a/docs/okn/semantic-review.md +++ b/docs/okn/semantic-review.md @@ -2739,7 +2739,7 @@ Source-declared metadata (typed dates retained; editorial values are not semanti Navigation/collection document role; links organize documents. -Source types: ['index']; topic=False. Source SHA-256: `ab80cb6a99b0b89611cce1915c2316ff02b151359261af6afb55a5f29139e90b`. +Source types: ['index']; topic=False. Source SHA-256: `64f4b7c5d7fea4e92a00f7677994c91f4d837cce5902c2361ed01874fdf10da5`. Source-declared metadata (typed dates retained; editorial values are not semantic confidence): diff --git a/docs/okn/statistics.md b/docs/okn/statistics.md index 5eddc9a4..455ddf66 100644 --- a/docs/okn/statistics.md +++ b/docs/okn/statistics.md @@ -12,11 +12,11 @@ pages, then regenerate; never edit this file. | Quantity | Count | |---|---| -| Pages (all described as `schema:CreativeWork`) | 2167 | -| Separate page-topic concepts (`skos:Concept`) | 2006 | +| Pages (all described as `schema:CreativeWork`) | 2170 | +| Separate page-topic concepts (`skos:Concept`) | 2009 | | Document-scoped source relation assertions | 555 | | Additional local keyword concepts | 1945 | -| Concept-valued keyword links | 7207 | +| Concept-valued keyword links | 7213 | | Navigation pages (`schema:CollectionPage`) | 73 | | Redirect pages | 4 | | Documents whose topics have reviewed external mappings | 47 | @@ -24,13 +24,13 @@ pages, then regenerate; never edit this file. | Reviewed exact matches | 40 | | Reviewed close matches | 6 | | Reviewed associative matches | 4 | -| Typed edges from `semantic_relations` | 10764 | -| Untyped edges from body links (`dcterms:references`) | 23908 | -| Triples | 79488 | +| Typed edges from `semantic_relations` | 10770 | +| Untyped edges from body links (`dcterms:references`) | 23924 | +| Triples | 79558 | | Links to concepts that have no page yet (not exported) | 0 | | Links to files outside the knowledge base (not exported) | 725 | | Self-links dropped | 71 | -| Latest page update | 2026-10-12 | +| Latest page update | 2026-11-15 | ## Pages by domain @@ -38,7 +38,7 @@ pages, then regenerate; never edit this file. |---|---| | `BioFirm` | 10 | | `agents` | 29 | -| `biology` | 193 | +| `biology` | 196 | | `citations` | 39 | | `cognitive` | 851 | | `free_energy_principle` | 58 | @@ -52,7 +52,7 @@ pages, then regenerate; never edit this file. | Type | Pages | |---|---| -| `concept` | 1334 | +| `concept` | 1337 | | `mathematical_concept` | 522 | | `knowledge_base` | 56 | | `citation` | 34 | @@ -95,7 +95,7 @@ pages, then regenerate; never edit this file. | Relation | Edges | |---|---| -| `relates` | 8043 | +| `relates` | 8049 | | `foundation` | 509 | | `implements` | 461 | | `related` | 442 | diff --git a/knowledge_base/biology/README.md b/knowledge_base/biology/README.md index b106bb85..5e7b3bb5 100644 --- a/knowledge_base/biology/README.md +++ b/knowledge_base/biology/README.md @@ -208,3 +208,9 @@ The [[AGENTS|biological agents documentation]] provides specific guidance for im - [[bee_communication_systems|Bee Communication Systems]] - [[comparative_social_insect_analysis|Comparative Analysis of Ant and Bee Social Systems]] - [[implementation_examples_social_insects|Implementation Examples Using Social Insect Code]] + +## Experiment-design reference cards + +- [[cohort_census_design]] — Inclusive censuses and survival-conditioned endpoint sampling. +- [[multilag_stratified_preparations]] — Testing coarse-state predictive closure across preparation histories and lags. +- [[cue_reliability_offspring_interventions]] — Separating information quality from reproductive consequences in evolutionary experiments. diff --git a/knowledge_base/biology/catalog.md b/knowledge_base/biology/catalog.md index 2ee88d9a..83d7527f 100644 --- a/knowledge_base/biology/catalog.md +++ b/knowledge_base/biology/catalog.md @@ -6,7 +6,7 @@ status: generated # Catalog: biology -All 190 pages under `knowledge_base/biology/`, with the opening line of each. Generated by `code/scripts/kb_index.py` — regenerate, do not edit. Back to the [[knowledge_base/biology/README|biology overview]]. +All 193 pages under `knowledge_base/biology/`, with the opening line of each. Generated by `code/scripts/kb_index.py` — regenerate, do not edit. Back to the [[knowledge_base/biology/README|biology overview]]. ## Top level @@ -52,6 +52,7 @@ All 190 pages under `knowledge_base/biology/`, with the opening line of each. Ge - [[knowledge_base/biology/chemical_ecology|Chemical Ecology]] — Chemical ecology studies how organisms use chemical signals and cues to mediate interactions with each other and with their environment. - [[knowledge_base/biology/chemogenetics|Chemogenetics]] — Methods that engineer receptors or channels to respond to otherwise inert small molecules, enabling remote control of defined cell populations, with designer receptors (DREADDs)… - [[knowledge_base/biology/climate_change_ecology|Climate Change Ecology]] — Climate change ecology studies how changes in temperature, precipitation, atmospheric carbon dioxide, and extreme events alter the physiology, distribution, timing, and… +- [[knowledge_base/biology/cohort_census_design|Cohort Census Design]] — Inclusive censuses and survival-conditioned endpoint sampling. - [[knowledge_base/biology/collective_migration|Collective Migration]] — The coordinated movement of groups of cells that keep their contacts, with examples from development, wound healing and cancer, and parallels to collective motion in animal groups. - [[knowledge_base/biology/collective_transport|Collective Transport]] — Cooperative carrying of loads too large for one individual, as seen in ants and implemented in groups of robots, and the coordination mechanisms and control problems it involves. - [[knowledge_base/biology/collision_avoidance|Collision Avoidance]] — How moving animals, groups and robots steer to avoid hitting neighbours or obstacles, covering separation rules, looming detection and social-force models, and how avoidance can… @@ -60,6 +61,7 @@ All 190 pages under `knowledge_base/biology/`, with the opening line of each. Ge - [[knowledge_base/biology/confusion_effect|Confusion Effect]] — The reduction in a predator's capture success when it attacks a group of similar moving prey, attributed to limits on how many targets the predator can track and single out. - [[knowledge_base/biology/conservation_biology_social_insects|Conservation Biology of Social Insects]] — The conservation of social insects, particularly ants and bees, is critical for maintaining biodiversity, ecosystem services, and ecological stability. - [[knowledge_base/biology/cortisol|Cortisol]] — Cortisol as the main human glucocorticoid, released through the HPA axis in response to stress and circadian cues, with its physiological effects and an allostatic Active… +- [[knowledge_base/biology/cue_reliability_offspring_interventions|Cue Reliability and Offspring Interventions]] — Separating information quality from reproductive consequences in evolutionary experiments. - [[knowledge_base/biology/cultural_services|Cultural Services]] — The non-material benefits that people obtain from ecosystems, such as recreation, spiritual and aesthetic value, education and heritage, and how they are assessed within the… - [[knowledge_base/biology/cuticular_hydrocarbons|Cuticular Hydrocarbons]] — Waxy long-chain hydrocarbons on the insect cuticle that limit water loss and, in social insects, act as chemical signatures used for nestmate recognition and caste signalling. - [[knowledge_base/biology/cytoskeleton|Cytoskeleton]] — The network of protein filaments inside eukaryotic cells, composed of actin filaments, intermediate filaments and microtubules, that supports cell shape, movement, division and… @@ -133,6 +135,7 @@ All 190 pages under `knowledge_base/biology/`, with the opening line of each. Ge - [[knowledge_base/biology/morphogenesis|Morphogenesis Through the Free Energy Principle]] — Morphogenesis -- the emergence of biological form and pattern during development -- is one of the most striking examples of self-organization in nature. - [[knowledge_base/biology/morphogenetic_fields|Morphogenetic Fields]] — The developmental-biology concept of spatial regions or gradients that coordinate cell fates during form generation, distinguished from the contested morphic fields of Sheldrake… - [[knowledge_base/biology/mrna_transport|mRNA Transport]] — The targeting of messenger RNA molecules to specific cellular locations, including neuronal dendrites and synapses, enabling local protein synthesis that supports synaptic… +- [[knowledge_base/biology/multilag_stratified_preparations|Multi-lag Stratified Preparations]] — Testing coarse-state predictive closure across preparation histories and lags. - [[knowledge_base/biology/mutual_aid|Mutual Aid]] — Cooperation among members of a species or group to their shared benefit, from Kropotkin's argument against pure competition to the modern theories of kin selection, reciprocity… - [[knowledge_base/biology/myelination|Myelination]] — Myelination as the wrapping of axons by oligodendrocytes and Schwann cells, which speeds conduction, follows a protracted developmental timeline, and is shaped by experience. - [[knowledge_base/biology/myrmecology|Myrmecology]] — Myrmecology is the scientific study of ants (family Formicidae), focusing on their biology, behavior, ecology, evolution, and taxonomy. diff --git a/knowledge_base/biology/cohort_census_design.md b/knowledge_base/biology/cohort_census_design.md new file mode 100644 index 00000000..d10dcb9c --- /dev/null +++ b/knowledge_base/biology/cohort_census_design.md @@ -0,0 +1,46 @@ +--- +title: Cohort Census Design +description: Inclusive censuses and survival-conditioned endpoint sampling. +type: concept +status: draft +created: 2026-11-15 +updated: 2026-11-15 +tags: + - evolution + - experimental_design +semantic_relations: + - type: relates + links: + - "[[natural_selection]]" + - "[[population_genetics]]" +--- + +# Cohort Census Design + +## Overview +A cohort census defines which individuals and reproductive events enter an analysis. Record the inclusion rule before estimating selection, drift, or transmission. An inclusive cohort retains zero-offspring parents and extinct lineages. An endpoint cohort conditioned on survival estimates a different distribution. + +## Inclusive versus endpoint-conditioned cohorts +Let K be descendant count and S the event that a lineage survives to the endpoint. The inclusive law is P(K); the selected law is P(K | S) = P(K,S)/P(S), when P(S)>0. Replacing one with the other changes the estimand. Record losses, censoring and missing parentage separately from observed zero reproduction; a missing record is not a zero. + +For an inclusive haploid census with N descendants sampled independently conditional on parental type frequency p, K | p is Binomial(N,p). If p varies across a shared environmental history, the unconditional law is a mixture. Its variance is N E[p(1-p)] + N² Var(p), not generally N p_bar(1-p_bar). Endpoint selection needs its own conditional likelihood; do not fit the inclusive binomial to selected survivors. + +## Experiment-design reference card +- Unit: a declared ancestral cohort and its assigned descendants, with explicit ploidy and census times. +- Preparation: replicate matched initial frequencies; record shared environmental history and family identity. +- Measurements: linked parentage, reproduction including zeros, survival, censoring, and inclusion probabilities. +- Comparison: report inclusive and endpoint-conditioned estimates separately, with the conditioning event in each likelihood. +- Limitation: the Price identity accounts for realized change; causal selection requires a design that separates correlated causes. + +## Relation to Active Inference +A survival-conditioned posterior and an inclusive population census have different conditioning information. A shared normalization formula alone does not equate their biological meaning. + +## Related concepts +- [[genetic_drift]] +- [[natural_selection]] +- [[population_genetics]] + +## References +- Price, G. R. (1970). Selection and covariance. Nature, 227, 520–521. https://doi.org/10.1038/227520a0 + +The sampling and variance formulas above follow directly from conditional probability and the law of total variance; this card proposes a design rather than reporting an experiment. diff --git a/knowledge_base/biology/cue_reliability_offspring_interventions.md b/knowledge_base/biology/cue_reliability_offspring_interventions.md new file mode 100644 index 00000000..00cd6496 --- /dev/null +++ b/knowledge_base/biology/cue_reliability_offspring_interventions.md @@ -0,0 +1,48 @@ +--- +title: Cue Reliability and Offspring Interventions +description: Separating information quality from reproductive consequences in evolutionary experiments. +type: concept +status: draft +created: 2026-11-15 +updated: 2026-11-15 +tags: + - evolution + - experimental_design +semantic_relations: + - type: relates + links: + - "[[natural_selection]]" + - "[[population_genetics]]" +--- + +# Cue Reliability and Offspring Interventions + +## Overview +A cue intervention changes information available before a choice. An offspring intervention changes reproductive consequences. Separate these interventions to test whether a fitted observation model predicts behavior and whether an action-value model predicts reproductive success. + +## Experiment-design reference card +- Declare environment e, cue c, action or expressed phenotype a, and descendant count at a specified inclusive census. +- Cue arm: vary P(c | e) while holding environmental frequencies, resources, action costs and offspring consequences fixed as far as the design permits. +- Consequence arm: change offspring consequences conditional on environment and action while holding cue statistics fixed. Freeze the controller during a short-term test; study adaptation in a separately identified arm. +- Factorial arm: cross cue reliability with consequence changes; record interactions rather than attributing every behavioral shift to sensing. +- Controls: cue-shuffled and cue-blind comparators, matched intervention costs, held-out choices, and independently measured offspring counts including zeros. +- Temporal controls: match environmental frequencies while varying their ordering; distinguish private cues from a shared environmental history. + +## Growth and inference +For independent environmental states with probabilities p(e), allocation pi(a), and nonnegative growth factors f(a,e), a simple large-population benchmark is sum_e p(e) log(sum_a pi(a) f(a,e)), where each inner sum is positive. A finite-horizon count, survival-conditioned endpoint, and long-run log-growth rate are different outcomes. State which is measured. + +In the special horse-race case, an action reproduces only in its matching environment and unconstrained allocation maximizes expected log growth by matching environmental probabilities, or their cue-conditioned probabilities when a cue is available. This model-specific result does not identify expected free energy with offspring count. Serial dependence, inherited state, density regulation and finite-population extinction need additional models. + +## Relation to Active Inference +Evaluate the observation mapping with cue interventions and evaluate the proposed link between action value and reproductive value with consequence interventions. Success on one test does not establish the other link. + +## Related concepts +- [[phenotypic_plasticity]] +- [[natural_selection]] +- [[cohort_census_design]] +- [[multilag_stratified_preparations]] + +## References +- Kelly, J. L., Jr. (1956). A new interpretation of information rate. Bell System Technical Journal, 35, 917–926. https://doi.org/10.1002/j.1538-7305.1956.tb03809.x + +The factorial protocol is a proposed diagnostic design; no empirical effect size or result is asserted. diff --git a/knowledge_base/biology/multilag_stratified_preparations.md b/knowledge_base/biology/multilag_stratified_preparations.md new file mode 100644 index 00000000..3f767e59 --- /dev/null +++ b/knowledge_base/biology/multilag_stratified_preparations.md @@ -0,0 +1,42 @@ +--- +title: Multi-lag Stratified Preparations +description: Testing coarse-state predictive closure across preparation histories and lags. +type: concept +status: draft +created: 2026-11-15 +updated: 2026-11-15 +tags: + - evolution + - experimental_design +semantic_relations: + - type: relates + links: + - "[[natural_selection]]" + - "[[population_genetics]]" +--- + +# Multi-lag Stratified Preparations + +## Overview +A coarse observable can hide predictive differences among fine states. Test closure using several preparations within the same coarse state and several observation lags. Agreement at one lag or in one stationary preparation is insufficient to establish autonomous coarse dynamics. + +## Experiment-design reference card +- Declare a fine-state variable X, coarse observation Y = g(X), preparation protocol, sampling clock, and observation error model. +- Within each starting coarse state, prepare distinct fine-state mixtures. Stratify the analysis by preparation history rather than pooling those mixtures immediately. +- Measure coarse transition distributions at multiple predeclared lags, using comparable measurement conditions and held-out preparations. +- Compare P(Y at t+lag | Y at t, preparation) across preparations at each lag. Differences expose predictive dependence on unresolved state or history. +- Include a fine-state or history-dependent predictor as a comparator. Match sample size and uncertainty before interpreting a null difference. + +## Finite-state criterion +For a row-stochastic fine transition matrix P, strong lumpability requires sum of P[x,x_next] over each destination block to be equal for all x in the same source block. Under this condition a coarse one-step transition Q exists for every initial preparation, and its powers predict all integer lags. Observed agreement at finitely many lags supports only those tested settings; it does not prove the universal criterion. Stationarity alone does not impose the block-sum equality. + +## Relation to Active Inference +A coarse generative model can predict well within one preparation while failing after a history intervention. Evaluate predictive adequacy separately from claims of exact closure or mechanistic identity. + +## Related concepts +- [[population_genetics]] +- [[evolutionary_dynamics]] +- [[cohort_census_design]] + +## References +The block-sum criterion is stated explicitly here as a finite-matrix condition. The preparation protocol is a proposed diagnostic design, with no claimed experimental result. diff --git a/knowledge_base/catalog.md b/knowledge_base/catalog.md index 12d7d35f..e0597671 100644 --- a/knowledge_base/catalog.md +++ b/knowledge_base/catalog.md @@ -12,7 +12,7 @@ Entry point to every domain catalog. Generated by `code/scripts/kb_index.py` — - [[knowledge_base/agents/catalog|agents]] — 16 pages - [[knowledge_base/BioFirm/catalog|BioFirm]] — 7 pages -- [[knowledge_base/biology/catalog|biology]] — 190 pages +- [[knowledge_base/biology/catalog|biology]] — 193 pages - [[knowledge_base/citations/catalog|citations]] — 36 pages - [[knowledge_base/cognitive/catalog|cognitive]] — 848 pages - [[knowledge_base/free_energy_principle/catalog|free energy principle]] — 41 pages