phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | machine generation of creative visual art— “The goal of this paper is to investigate a computational creative system for art generation without involving a human artist in the creative process, but nevertheless involving human creative products in the learning process.” | |
| claim-type | method— “We propose a new system for generating art.” | |
| method — evidence | — “We propose a model for an art-generating agent, and then propose a functioning model using a variant of GAN to make it creative.” | |
| transfer | from artificial intelligence (deep learning / GANs): generative adversarial networks adapted to generate art that deviates from style norms— “We build over Generative Adversarial Networks (GAN), which have shown the ability to learn to generate novel images simulating a given distribution” | transfers[0] |
| transfer | from psychology (Berlyne aesthetics / arousal theory): Berlyne theory of aesthetic novelty and arousal used to design the creativity objective— “Berlyne emphasized that the most significant arousal-raising properties for aesthetics are novelty” | transfers[1] |
| mechanism | modify the GAN objective to push generated images to deviate from established style norms while staying near the art distribution— “force the system to explore creative space to find solutions that deviate from established styles but stay close enough to the boundary” | mechanisms[0] |
| extends | Generative Adversarial Networks (from AI)— “We build over Generative Adversarial Networks (GAN)” | claim_relations[0] |
| builds-on | Generative Adversarial Networks (GAN)— “We build over Generative Adversarial Networks (GAN), which have shown the ability to learn to generate novel images simulating a given distribution.” | claim_relations[1] |
| extends | Generative Adversarial Networks (GAN)— “We propose modifications to its objective to make it capable of generating creative art by maximizing deviation from established styles and minimizing deviation from art distribution.” | claim_relations[2] |
| builds-on | Martindale's principle of least effort and explanation of style breaks— “The approach proposed in this paper is inspired by Martindale’s principle of least effort and his explanation of style breaks.” | claim_relations[3] |
| builds-on | the art-perception experiment of Snapper et al [23]— “This experiment is similar to an experiment conducted by Snapper et al [23] to determine to what degree human subjects find the works of art to be intentional, having visual structure, communicative, and inspirational.” | claim_relations[4] |
| sense | style - AS A CLASS IN A CLASSIFIER: the network's final layer has K outputs where K is the number of style classes, so here the word names a position in a label set rather than any property of the picture, and an image's style is whichever index wins— “into 3 fully collected layers sizes 1024, 512, K, respectively, where K is the number of style classes” | senses[0] |
| sense | style ambiguity - AS POSTERIOR ENTROPY: the generator is rewarded for making the style classifier uncertain, so 'deviating from style norms' is implemented as producing images the classifier cannot place, and creativity becomes a property of a confusion matrix— “We added a style classification loss and a style ambiguity loss. Maximizing the stylistic ambiguity can be achieved by maximizing the style class posterior entropy” | senses[1] |
| sense | novelty - INHERITED FROM BERLYNE'S COLLATIVE VARIABLES: it is one of five arousal-raising properties named in an aesthetics theory - novelty, surprisingness, complexity, ambiguity and puzzlingness - and this system implements the fourth of them while the record's limitation says creativity was operationalised as the first— “Berlyne emphasized that the most significant arousal-raising properties for aesthetics are novelty, surprisingness, complexity, ambiguity, and puzzlingness. He coined the term collative variables to refer to these properties collectively” | senses[2] |
| control | an art-trained DCGAN baseline at both 64x64 and 256x256, generated and rated alongside the proposed model by the same human subjects — held — rules out the human ratings reflecting that the images are art-like at all rather than that the proposed objective did anything - the baseline is the same architecture without it— “Top Panel: Sample images generated by art-trained DC- GAN 64x64 baseline model (top 6 and bottom 6 ranked images are shown according to human subject likeness rating in Experiment I).” | controls[0] |
| audit | style ambiguity —operationalised-as→ style class — Audit 2026-09-28: DOUBTFUL - read against its full span and kept, but ambiguity is achieved by maximizing the style-class posterior entropy; operationalised-as is loose.— “We added a style classification loss and a style ambiguity loss. Maximizing the stylistic ambiguity can be achieved by maximizing the style class posterior entropy” | concept_edges[0] |
| edge | novelty —part-of→ collative variables— “Berlyne emphasized that the most significant arousal-raising properties for aesthetics are novelty, surprisingness, complexity, ambiguity, and puzzlingness. He coined the term collative variables to refer to these properties collectively” | concept_edges[1] |
| outcome (stated by the tagger, no quotation) | control — held | controls[0] |
| for which use | limitation — the generative approach itself: THE FUNDAMENTAL LIMITATION OF USING GANS IN GENERATING CREATIVE WORKS— “the fundamental limitation of using GANs in generating creative works” | limitations[0] |
| for which use | limitation — every creativity claim in the paper: ASSESSING THE CREATIVITY OF ARTIFACTS GENERATED BY THE MACHINE IS AN OPEN AND HARD QUESTION— “Assessing the creativity of artifacts generated by the machine is an open and hard question.” | limitations[1] |
| for which use | limitation — what the word creative means in this work: SINCE INFLUENCE IS NOT RELEVANT HERE, WE USE NOVELTY AS A PROXY FOR CREATIVITY— “However, since influence is not relevant here, we use novelty as a proxy for creativity.” | limitations[2] |
| for which use | limitation — any semantic reading of what the system does: IT DOES NOT HAVE ANY SEMANTIC UNDERSTANDING OF ART BEHIND THE CONCEPT OF STYLES— “However it does not have any semantic understanding of art behind the concept of styles.” | limitations[3] |
| for which use | limitation — the headline comparison, which the authors decline to interpret: WE LEAVE OPEN HOW TO INTERPRET THE HUMAN SUBJECTS' RESPONSES THAT RANKED THE CAN ART BETTER THAN THE ART BASEL SAMPLES— “We leave open how to interpret the human subjects’ responses that ranked the CAN art better than the Art Basel samples in different aspects.” | limitations[4] |
| for which use | limitation — the same scope as limitations[4]: MORE EXPERIMENTS ARE DEFINITELY NEEDED TO HELP ANSWER THESE QUESTIONS— “More experiments are definitely needed to help answer these questions.” | limitations[5] |
| for which use | limitation — the creativity claimed for these images: it is implemented as maximising the entropy of the style class posterior, that is, as being hard for the style classifier to assign to any class— “Maximizing the stylistic ambiguity can be achieved by maximizing the style class posterior entropy” | limitations[6] |
| for which use | limitation — the style categories used here: they are a fixed set whose size K is a network design parameter, fixed by the label set of the training corpus rather than by art history— “into 3 fully collected layers sizes 1024, 512, K, respectively, where K is the number of style classes” | limitations[7] |
| for which use | limitation — the claim that this system generates art: the paper's title puts the word in quotation marks and describes the method as learning about styles and deviating from style norms— “CAN: Creative Adversarial Networks Generating “Art” by Learning About Styles and Deviating from Style Norms” | limitations[8] |
| replication | untested | |
| limitation | GANs have a fundamental limitation in generating genuinely creative works— “the fundamental limitation of using GANs in generating creative works” | limitations[0] |
| limitation | Creativity of machine-generated artifacts is acknowledged to be an unresolved assessment problem.— “Assessing the creativity of artifacts generated by the machine is an open and hard question.” | limitations[1] |
| limitation | Creativity was operationalised only as novelty, because influence could not be measured.— “However, since influence is not relevant here, we use novelty as a proxy for creativity.” | limitations[2] |
| limitation | The system has no semantic understanding of art beyond style categories.— “However it does not have any semantic understanding of art behind the concept of styles.” | limitations[3] |
| limitation | The authors cannot explain why subjects rated CAN images above the human Art Basel set.— “We leave open how to interpret the human subjects’ responses that ranked the CAN art better than the Art Basel samples in different aspects.” | limitations[4] |
| limitation | Further experiments are needed to settle the interpretation of the results.— “More experiments are definitely needed to help answer these questions.” | limitations[5] |
| limitation | Creativity is implemented as making a classifier maximally unsure which style it is looking at.— “Maximizing the stylistic ambiguity can be achieved by maximizing the style class posterior entropy” | limitations[6] |
| limitation | The style vocabulary is a fixed list whose length is a design parameter of the network.— “into 3 fully collected layers sizes 1024, 512, K, respectively, where K is the number of style classes” | limitations[7] |
| limitation | The paper's own object is scare-quoted in its title. | limitations[8] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | CAN (Creative Adversarial Networks) generates art by learning styles, then becoming creative by increasing arousal potential — maximizing deviation from established styles while minimizing deviation from the overall art distribution. | content_tags.findings[0] |
| not machine-vouched | finding — declared-ellipsis - An elided quote: its 2 pieces ('The system generates art by look- ing at' ... 'izing de- viation from art distribution.') are each in our copy of the paper (with line-break hyphens, spacing and punctuation set aside), and the quote marks the cut with an ellipsis, so the joined string is in no copy. Located mechanically, piece by piece. | content_tags.findings[0] |
| finding | Experiments compared human subjects' responses to the generated art with their responses to art created by people. | content_tags.findings[1] |
| variable | arousal potential (independent) | content_tags.variables[0] |
| definition | Creative Adversarial Network (CAN) | content_tags.definitions[0] |
| evidence status note | definition — QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. | content_tags.definitions[0] |
| definition | collative variables | content_tags.definitions[1] |
| definition | Wundt curve Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.definitions[2] |
| key-terms | art, style, creative, GAN, arousal, novelty, generated, discriminator, deviating |
| concept | style Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[2] |
| concept | style class Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[4] |
| concept | style ambiguity Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[5] |
| concept | collative variables Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[6] |
| concept | novelty Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | concepts[7] |
| hypothesis-testing | as written: hypothesis Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[0] |
| evidence status note | hypothesis-testing — QUOTE REPAIRED: the extractor split a unicode and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. | content_tags.statistics[0] |
| null-hypothesis H0 | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[1] |
| evidence status note | null-hypothesis — QUOTE REPAIRED: the extractor split a unicode and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. | content_tags.statistics[1] |
| p-value p | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[2] |
| standard-deviation SD/σ | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[3] |
| evidence status note | standard-deviation — QUOTE REPAIRED: the extractor split a hyphen-space and the quote was recorded faithfully with the damage in it, which made it unanchorable. The repaired string is an EXACT SUBSTRING of the staged text - that is why it was accepted, not because a rule produced it. The text as originally recorded is kept in evidence as extracted. | content_tags.statistics[3] |
| t-test t | Quotation withheld — this paper is not ours to reproduce, and the page has already quoted as much of it as the rights layer allows. The tag, its field and its confidence stand; the sentence is at the publisher. | content_tags.statistics[4] |
When we have changed our mind. Some entries in the table begin with a word in capitals and the reason for it. EVIDENCE CORRECTED — the finding still stands, but we had quoted the wrong sentence from the paper, and the quotation has been replaced with the right one. RETRACTED — the entry itself was wrong; it stays on the page, marked, with the reason, rather than being deleted. WITHDRAWN AND REPLACED — the entry was wrong, and a corrected version of it appears in the entries below. Each mark carries its own reason, in brackets, because the reasons differ from one entry to the next. An entry we have WITHDRAWN is removed from this page and kept in the record, with its reason. A line reading Audit and a date is a later re-reading of the entry against its whole sentence: DOUBTFUL means it was kept, and the reason says what is loose about it. Quotation marks inside the tagger's sentence. Where a tag is the tagger's own sentence, a phrase in quotation marks inside it is borrowed from the paper — the tagger's wording around it is not. The paper's own sentence is the evidence quote for that row; on this page it is withheld with the article, and the article is at its publisher. Nothing inside quotation marks anywhere on this page is ours: if it is quoted, it is the paper's. Whose words a concept carries. A concept entry records a term the paper itself uses, so the term, the sense given for it, and its evidence are the paper's own words, quoted where the paper fixes them. This is the opposite of a sense entry in the tag table, which is the tagger's sentence about how the paper uses a term. Where a concept entry could not be grounded in the paper's words, the entry says so on its own row.
Tag layer © the Meta-Tagging Project, CC BY 4.0. The paper text is © its authors/publisher
and is linked, not reproduced. Excerpts shown are short quotations for identification and verification.
· back to index
· concepts · authors
· עברית
© 2026 Shir Sivroni · ORCID 0009-0000-2597-2824