https://w3id.org/np/RANwhBVuI2oACSeyoj_wscMMY62zERpjvWqGVW9Yc0lcE
.trig | .trig.txt | .jelly | .jelly.txt | .jsonld | .jsonld.txt | .nq | .nq.txt | .xml | .xml.txt
@prefix this: <https://w3id.org/np/RANwhBVuI2oACSeyoj_wscMMY62zERpjvWqGVW9Yc0lcE> .
@prefix sub: <https://w3id.org/np/RANwhBVuI2oACSeyoj_wscMMY62zERpjvWqGVW9Yc0lcE/> .
@prefix np: <http://www.nanopub.org/nschema#> .
@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
@prefix prov: <http://www.w3.org/ns/prov#> .
@prefix npx: <http://purl.org/nanopub/x/> .
@prefix dc: <http://purl.org/dc/terms/> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
sub:Head {
this: a np:Nanopublication;
np:hasAssertion sub:assertion;
np:hasProvenance sub:provenance;
np:hasPublicationInfo sub:pubinfo .
}
sub:assertion {
<https://w3id.org/delphi/resource/assessment/levin-2023-competency-09--claim-test>
a <https://w3id.org/delphi/ontology/Assessment>;
dc:description "analysis.py ran the evolvable-competency genetic algorithm for 3000 generations, 20 independent runs at each of penalties 0, 1e-7, 1e-4 and 0.5. Selection score was raw phenotypic fitness minus penalty times the competency gene, the subtraction named in the figure 7 caption. The tracked embryo each generation was the lowest-index maximizer of that score. Early window means used generations 21-100 and late window means used generations 2701-3000. The pre-registered rule was then applied to the genotypic deltas, the competency deltas, the 15-of-20 counts, phenotypic maintenance, and the order of the three penalized genotypic deltas.";
<http://www.w3.org/2000/01/rdf-schema#comment> "Measured: 0.411819 mean late-minus-early structural-gene fitness of the selected embryo at penalty 1e-4 (late generations 2701-3000 minus early generations 21-100) (n = 20, fixed seed 2026110209; 20-run sign counts are in extra). Interpretation: Simulation-result claim. The full results paragraph says penalty factors from 1e-7 to 0.5 produced a Baldwin effect: the rate of rise of genotypic fitness rose with the penalty, structural-gene fitness kept improving, and selection preferred progressively lower competency, while phenotypic fitness stayed at its maximum. For factors over 0.5 the effect disappeared; 0.5 itself is inside the range they say worked. Figure 7 uses 1e-4 times the competency value over 3000 generations and n = 100; this test uses that subtraction, that duration, and n = 20. The released repository does not contain the penalty experiment. The run shows whether this program behaves as claimed. It does not show that cells pay a competency cost or that the Baldwin effect occurs in organisms. Real competency costs are unmeasured, as the authors note, and that sentence is not tested.";
<http://www.w3.org/2000/01/rdf-schema#label> "Claim test of levin-2023-competency-09: mixed, confidence 0.73";
<http://www.w3.org/2000/01/rdf-schema#seeAlso> <https://github.com/fractastical/delphisci/blob/bf2af6ee0a84e23ffa6414ad7a0a50a8c276506e/test-data/claim-tests/levin-papers/levin-2023-competency-09/result.json>;
prov:wasAttributedTo <https://w3id.org/delphi/resource/agent/delphisci-claim-tester-grok-4-7>;
<https://w3id.org/delphi/ontology/admissibleDatasets> "Runs of the same evolvable-competency genetic algorithm as levin-2023-competency-12 (authors' released randint bounds, population 100, 3000 generations), with selection on phenotypic fitness minus penalty times competency. The released code contains no penalty experiment, so an independent extension that states the subtraction formula from the figure 7 caption is admissible. Biological measurements of competency cost are not admissible for this simulation-result claim.";
<https://w3id.org/delphi/ontology/assessmentType> "claim-test/0.1.0";
<https://w3id.org/delphi/ontology/dataKind> "computational";
<https://w3id.org/delphi/ontology/disconfirmingObservation> "At penalty 1e-4, structural-gene fitness does not rise from the early window to the late window, or competency does not fall.";
<https://w3id.org/delphi/ontology/evidenceAgainst> "At penalty 0.5, competency was already 1.0, the lowest value the mutation operator writes, in both windows, so it did not fall further (0 of 20 runs). Structural-gene fitness rose only from 0.933 to 0.999. The genotypic deltas are 0.066, 0.412 and 0.406 at penalties 0.5, 1e-4 and 1e-7, so the rise is not ordered with the penalty.",
"At penalty 1e-4, structural-gene fitness of the selected embryo rose from 0.560 in generations 21-100 to 0.971 in generations 2701-3000 (20 of 20 runs), competency fell from 348 to 18 (20 of 20), and raw phenotypic fitness in the late window was 1.0. Penalty 1e-7 did the same: fitness 0.565 to 0.971, competency 339 to 18, 20 of 20 runs both ways.",
"The mixed verdict is the pre-registered rule: the figure's penalty moves in both locked directions, and the top of the claimed range does not. Real competency costs are unmeasured, and this run does not show that cells pay one.",
"This is an independent reimplementation. The authors' repository has no penalty experiment. The subtraction formula is the figure 7 caption ('a fitness penalty of 1e-4 times the competency value'), not a line of their code.",
"With no penalty, structural-gene fitness fell slightly (delta -0.034, 1 of 20 runs rose) and competency rose from 446 to 487 (2 of 20 runs fell). The no-cost control does not show the Baldwin pair.";
<https://w3id.org/delphi/ontology/expectedObservation> "For every penalty in {1e-7, 1e-4, 0.5}, late structural-gene fitness of the selected embryo exceeds its early value and late competency is lower, phenotypic fitness in the late window stays at least 0.95, and larger penalties produce a larger genotypic rise. The same pair of directions does not both hold at penalty 0.";
<https://w3id.org/delphi/ontology/failCriterion> "Pre-registered: refutes if, at the figure's penalty 1e-4, the mean genotypic delta is <= 0 or the mean competency delta is >= 0. Mixed if 1e-4 moves in both locked directions but any other pass requirement fails, including an endpoint of the range, the 15/20 counts, phenotypic maintenance, the strict rate order, or a penalty-0 control that also meets both 15/20 direction rules.";
<https://w3id.org/delphi/ontology/hasConfidence> "0.73"^^xsd:double;
<https://w3id.org/delphi/ontology/hasEvidence> <https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-1>,
<https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-2>,
<https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-3>;
<https://w3id.org/delphi/ontology/hasVerdict> "mixed";
<https://w3id.org/delphi/ontology/hypothesis> "Penalizing phenotypic fitness by a factor times the competency value, for factors from 1e-7 through 0.5, produces a Baldwin effect in the stated direction: the best embryo's structural-gene fitness is higher late in a 3000-generation run than just after the first 20 generations, and the competency gene of that embryo is lower late than early.";
<https://w3id.org/delphi/ontology/inferenceMethod> "Window means of the selected embryo's structural-gene fitness and competency gene, early generations 21-100 versus late generations 2701-3000, 20 runs per penalty. Directions require a positive or negative mean delta and agreement in at least 15 of 20 runs. Rate order is the strict order of the three mean genotypic deltas.";
<https://w3id.org/delphi/ontology/passCriterion> "Pre-registered before any penalized run. Selection score = raw phenotypic fitness minus penalty times the competency gene. The best embryo each generation is the lowest-index maximizer of that score. Genotypic fitness is that embryo's structural-gene fitness. Competency is that embryo's competency gene. Raw phenotypic fitness is its unpenalized developed fitness. Early window: generations 21-100. Late window: generations 2701-3000. Penalties 1e-7, 1e-4 and 0.5, each with 20 independent runs of 3000 generations. A penalty passes its direction if all of: the mean over runs of (late-window mean genotypic fitness minus early-window mean) is > 0; at least 15 of 20 runs have late mean genotypic > early mean genotypic; the mean over runs of (late-window mean competency minus early-window mean) is < 0; at least 15 of 20 runs have late mean competency < early mean competency; the mean over runs of the late-window mean raw phenotypic fitness is >= 0.95. Supports only if every one of those three penalties passes, the three mean genotypic deltas are strictly ordered delta(0.5) > delta(1e-4) > delta(1e-7), and a penalty-0 control of 20 runs does not itself pass both the genotypic 15/20 rule and the competency 15/20 rule.";
<https://w3id.org/delphi/ontology/preregistered> "false"^^xsd:boolean .
<https://w3id.org/delphi/resource/claim/levin-2023-competency-09> <https://w3id.org/delphi/ontology/hasAssessment>
<https://w3id.org/delphi/resource/assessment/levin-2023-competency-09--claim-test> .
<https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-1>
a <https://w3id.org/delphi/ontology/EvidenceItem>;
dc:description "window means of structural-gene fitness, competency and raw phenotypic fitness; summary in derived/measured.json";
dc:license "";
<http://www.w3.org/2000/01/rdf-schema#label> "Dataset: Penalty sweep of the evolvable-competency model, 20 runs x 3000 generations at penalties 0, 1e-7, 1e-4 and 0.5 (generated by analysis.py)" .
<https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-2>
a <https://w3id.org/delphi/ontology/EvidenceItem>;
dc:description "section 3.5 and the figure 7 caption, which states a fitness penalty of 1e-4 times the competency value";
dc:license "CC BY 4.0";
<http://www.w3.org/2000/01/rdf-schema#label> "Dataset: Shreesha L, Levin M (2023) Cellular Competency during Development Alters Evolutionary Dynamics in an Artificial Embryogeny Model. Entropy 25(1):131";
prov:wasDerivedFrom <https://doi.org/10.3390/e25010131> .
<https://w3id.org/delphi/resource/evidence/levin-2023-competency-09--claim-test--dataset-3>
a <https://w3id.org/delphi/ontology/EvidenceItem>;
dc:description "evolvable-competency genetic algorithm; the penalty experiment is not in this repository";
dc:license "";
<http://www.w3.org/2000/01/rdf-schema#label> "Dataset: Authors' simulation code for Shreesha and Levin (2023)";
prov:atLocation <https://raw.githubusercontent.com/Niwhskal/CellularCompetency/2be29f96628c83613074e1d2b874e66cff9279cc/all_tests/Exp3-EvolvableCompetency.py>,
<https://raw.githubusercontent.com/Niwhskal/CellularCompetency/2be29f96628c83613074e1d2b874e66cff9279cc/all_tests/core_functions.py>,
<https://raw.githubusercontent.com/Niwhskal/CellularCompetency/2be29f96628c83613074e1d2b874e66cff9279cc/hyperparameters.json>;
prov:wasDerivedFrom <https://github.com/Niwhskal/CellularCompetency>;
<https://w3id.org/delphi/ontology/hasChecksum> "sha256:55ad59da900e896f2b59b153c8dcdfe9371d0c3f01a8fc5ba3638a430badc1d6",
"sha256:8357fe689efe9ab76efd8c5f39bbf474a5b15ff4381be856b17a6dbc7c080c46", "sha256:b172cf3b2d5858636156156730ba710e5972eba0c38f5a5460021d7717890ec9" .
}
sub:provenance {
<https://w3id.org/delphi/resource/agent/delphisci-claim-tester-grok-4-7> a prov:SoftwareAgent;
<http://www.w3.org/2000/01/rdf-schema#label> "delphisci claim tester (grok-4.7; cursor/levin-competency-goals-363e)";
prov:actedOnBehalfOf <https://orcid.org/0009-0007-4948-9192>;
<https://w3id.org/delphi/ontology/usesModel> <https://w3id.org/delphi/resource/model/grok-4-7> .
<https://w3id.org/delphi/resource/model/grok-4-7> a <https://w3id.org/delphi/ontology/LanguageModel>;
dc:identifier "grok-4.7" .
sub:assertion prov:wasAttributedTo <https://w3id.org/delphi/resource/agent/delphisci-claim-tester-grok-4-7>;
prov:wasDerivedFrom <https://github.com/fractastical/delphisci/blob/bf2af6ee0a84e23ffa6414ad7a0a50a8c276506e/test-data/claim-tests/levin-papers/levin-2023-competency-09/result.json>;
prov:wasGeneratedBy sub:test .
sub:plan prov:hadPlan <https://github.com/fractastical/delphisci/blob/bf2af6ee0a84e23ffa6414ad7a0a50a8c276506e/test-data/claim-tests/README.md> .
sub:test a prov:Activity;
prov:qualifiedAssociation sub:plan;
prov:wasAssociatedWith <https://w3id.org/delphi/resource/agent/delphisci-claim-tester-grok-4-7> .
}
sub:pubinfo {
this: dc:created "2026-11-08T16:24:31.881Z"^^xsd:dateTime;
dc:creator <https://orcid.org/0009-0007-4948-9192>;
dc:license <https://creativecommons.org/licenses/by/4.0/>;
npx:hasNanopubType <https://w3id.org/delphi/ontology/Assessment>;
npx:introduces <https://w3id.org/delphi/resource/assessment/levin-2023-competency-09--claim-test>;
<http://www.w3.org/2000/01/rdf-schema#label> "Claim test of levin-2023-competency-09: mixed, confidence 0.73";
<https://w3id.org/delphi/ontology/domain> <https://w3id.org/delphi/resource/domain/levin-papers> .
sub:sig npx:hasAlgorithm "RSA";
npx:hasPublicKey "MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAyFHtPB6p0mCjDozxlOY9CmO3nBm7Te3jFbXyLmM+G1tGe47Da0jYIbhkl9Oo1ZjxmCa3LCwlEqK8gj6hEIcL4ahW30JXh9HycfKVk1bugkzMEV15tJdTzAU3sGMFnbNsn43xY2LFLy45tApkRm5byEiM7oRRrooIdwoQKnP4Wzq/bfAoJ9qvW+MGkelyOTd1jbSt+KfmT7xKmDfWzAHddUlWwMX2QJOLxDaffOteTop6ZRgPiNmPJKA/S62Gac9XtSYT3oDkZ6/nr5cxxNLpOyxTf5FPISiCe11xYO/unYjY+mwnOfA8yhJIyx1l0O9I0vpOpn+E5TvyHJV+oSTrvwIDAQAB";
npx:hasSignature "MIjoruH/E0UCV0CmFNwHqrGxXJ63QaOgjJu+RdP9AkQujYp8KPKkHtHgfxNGDTvewudTEQF0m89QQ/5g80qTW8mfr50QtBrSMVI8UmolZ2KZ3FxLhONihbE+0VnTRnTc//pXpn5lPs05tho6MX/Gm4GenO/25jSEmzRFNoPfzxi3ofbZm42Ll+LLyXltHRyq1+Q1LHT3UhCfoJnXBJEExI2hg01qxxzWy08nVolXahIw/zCK5i03WuUMEcxhBVAtJ1FwlzGd4cZa8tIu2Bsb+GMGdadsq06ZBnbbvlvYrJuOIOU2nh+hQjFJdPf86mVAg13HpHFfYb4pBVG1So2YcQ==";
npx:hasSignatureTarget this:;
npx:signedBy <https://orcid.org/0009-0007-4948-9192> .
}