Learning to Validate Generative Models: a Goodness-of-Fit Approach Article Swipe
Marco Letizia
,
Humberto Reyes-González
,
M. Zanetti
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2511.09118
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.48550/arxiv.2511.09118
Generative models are increasingly central to scientific workflows, yet their systematic use and interpretation require a proper understanding of their limitations through rigorous validation. Classic approaches struggle with scalability, statistical power, or interpretability when applied to high-dimensional data, making it difficult to certify the reliability of these models in realistic, high-dimensional scientific settings. Here, we propose the use of the New Physics Learning Machine (NPLM), a learning-based approach to goodness-of-fit testing inspired by the Neyman--Pearson construction, to test generative networks trained on high-dimensional scientific data. We demonstrate the performance of NPLM for validation in two benchmark cases: generative models trained on mixtures of Gaussian models with increasing dimensionality, and a public end-to-end model, known as FlowSim, developed to generate high-energy physics collision events. We demonstrate that the NPLM can serve as a powerful validation method while also providing a means to diagnose sub-optimally modeled regions of the data.
Related Topics
Concepts
No concepts available.
Metadata
- Type
- preprint
- Landing Page
- http://arxiv.org/abs/2511.09118
- https://arxiv.org/pdf/2511.09118
- OA Status
- green
- OpenAlex ID
- https://openalex.org/W4416716555
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4416716555Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.48550/arxiv.2511.09118Digital Object Identifier
- Title
-
Learning to Validate Generative Models: a Goodness-of-Fit ApproachWork title
- Type
-
preprintOpenAlex work type
- Publication year
-
2025Year of publication
- Publication date
-
2025-11-12Full publication date if available
- Authors
-
Marco Letizia, Humberto Reyes-González, M. ZanettiList of authors in order
- Landing page
-
https://arxiv.org/abs/2511.09118Publisher landing page
- PDF URL
-
https://arxiv.org/pdf/2511.09118Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
greenOpen access status per OpenAlex
- OA URL
-
https://arxiv.org/pdf/2511.09118Direct OA link when available
- Cited by
-
0Total citation count in OpenAlex
Full payload
| id | https://openalex.org/W4416716555 |
|---|---|
| doi | https://doi.org/10.48550/arxiv.2511.09118 |
| ids.doi | https://doi.org/10.48550/arxiv.2511.09118 |
| ids.openalex | https://openalex.org/W4416716555 |
| fwci | |
| type | preprint |
| title | Learning to Validate Generative Models: a Goodness-of-Fit Approach |
| biblio.issue | |
| biblio.volume | |
| biblio.last_page | |
| biblio.first_page | |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| language | |
| locations[0].id | pmh:oai:arXiv.org:2511.09118 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4306400194 |
| locations[0].source.issn | |
| locations[0].source.type | repository |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | arXiv (Cornell University) |
| locations[0].source.host_organization | https://openalex.org/I205783295 |
| locations[0].source.host_organization_name | Cornell University |
| locations[0].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[0].license | cc-by |
| locations[0].pdf_url | https://arxiv.org/pdf/2511.09118 |
| locations[0].version | submittedVersion |
| locations[0].raw_type | text |
| locations[0].license_id | https://openalex.org/licenses/cc-by |
| locations[0].is_accepted | False |
| locations[0].is_published | False |
| locations[0].raw_source_name | |
| locations[0].landing_page_url | http://arxiv.org/abs/2511.09118 |
| locations[1].id | doi:10.48550/arxiv.2511.09118 |
| locations[1].is_oa | True |
| locations[1].source.id | https://openalex.org/S4306400194 |
| locations[1].source.issn | |
| locations[1].source.type | repository |
| locations[1].source.is_oa | True |
| locations[1].source.issn_l | |
| locations[1].source.is_core | False |
| locations[1].source.is_in_doaj | False |
| locations[1].source.display_name | arXiv (Cornell University) |
| locations[1].source.host_organization | https://openalex.org/I205783295 |
| locations[1].source.host_organization_name | Cornell University |
| locations[1].source.host_organization_lineage | https://openalex.org/I205783295 |
| locations[1].license | cc-by |
| locations[1].pdf_url | |
| locations[1].version | |
| locations[1].raw_type | article |
| locations[1].license_id | https://openalex.org/licenses/cc-by |
| locations[1].is_accepted | False |
| locations[1].is_published | |
| locations[1].raw_source_name | |
| locations[1].landing_page_url | https://doi.org/10.48550/arxiv.2511.09118 |
| indexed_in | arxiv, datacite |
| authorships[0].author.id | https://openalex.org/A5075285734 |
| authorships[0].author.orcid | https://orcid.org/0000-0001-9641-4352 |
| authorships[0].author.display_name | Marco Letizia |
| authorships[0].author_position | last |
| authorships[0].raw_author_name | Letizia, Marco |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5033519909 |
| authorships[1].author.orcid | https://orcid.org/0000-0003-3283-5208 |
| authorships[1].author.display_name | Humberto Reyes-González |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Reyes-González, Humberto |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5100608249 |
| authorships[2].author.orcid | https://orcid.org/0000-0003-4281-4582 |
| authorships[2].author.display_name | M. Zanetti |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Zanetti, Marco |
| authorships[2].is_corresponding | False |
| has_content.pdf | True |
| has_content.grobid_xml | False |
| is_paratext | False |
| open_access.is_oa | True |
| open_access.oa_url | https://arxiv.org/pdf/2511.09118 |
| open_access.oa_status | green |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-11-14T00:00:00 |
| display_name | Learning to Validate Generative Models: a Goodness-of-Fit Approach |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-28T21:21:56.193364 |
| primary_topic | |
| cited_by_count | 0 |
| locations_count | 2 |
| best_oa_location.id | pmh:oai:arXiv.org:2511.09118 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4306400194 |
| best_oa_location.source.issn | |
| best_oa_location.source.type | repository |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | arXiv (Cornell University) |
| best_oa_location.source.host_organization | https://openalex.org/I205783295 |
| best_oa_location.source.host_organization_name | Cornell University |
| best_oa_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| best_oa_location.license | cc-by |
| best_oa_location.pdf_url | https://arxiv.org/pdf/2511.09118 |
| best_oa_location.version | submittedVersion |
| best_oa_location.raw_type | text |
| best_oa_location.license_id | https://openalex.org/licenses/cc-by |
| best_oa_location.is_accepted | False |
| best_oa_location.is_published | False |
| best_oa_location.raw_source_name | |
| best_oa_location.landing_page_url | http://arxiv.org/abs/2511.09118 |
| primary_location.id | pmh:oai:arXiv.org:2511.09118 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4306400194 |
| primary_location.source.issn | |
| primary_location.source.type | repository |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | arXiv (Cornell University) |
| primary_location.source.host_organization | https://openalex.org/I205783295 |
| primary_location.source.host_organization_name | Cornell University |
| primary_location.source.host_organization_lineage | https://openalex.org/I205783295 |
| primary_location.license | cc-by |
| primary_location.pdf_url | https://arxiv.org/pdf/2511.09118 |
| primary_location.version | submittedVersion |
| primary_location.raw_type | text |
| primary_location.license_id | https://openalex.org/licenses/cc-by |
| primary_location.is_accepted | False |
| primary_location.is_published | False |
| primary_location.raw_source_name | |
| primary_location.landing_page_url | http://arxiv.org/abs/2511.09118 |
| publication_date | 2025-11-12 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.a | 15, 65, 109, 131, 138 |
| abstract_inverted_index.We | 85, 123 |
| abstract_inverted_index.as | 114, 130 |
| abstract_inverted_index.by | 72 |
| abstract_inverted_index.in | 48, 93 |
| abstract_inverted_index.it | 39 |
| abstract_inverted_index.of | 18, 45, 58, 89, 102, 145 |
| abstract_inverted_index.on | 81, 100 |
| abstract_inverted_index.or | 31 |
| abstract_inverted_index.to | 5, 35, 41, 68, 76, 117, 140 |
| abstract_inverted_index.we | 54 |
| abstract_inverted_index.New | 60 |
| abstract_inverted_index.and | 12, 108 |
| abstract_inverted_index.are | 2 |
| abstract_inverted_index.can | 128 |
| abstract_inverted_index.for | 91 |
| abstract_inverted_index.the | 43, 56, 59, 73, 87, 126, 146 |
| abstract_inverted_index.two | 94 |
| abstract_inverted_index.use | 11, 57 |
| abstract_inverted_index.yet | 8 |
| abstract_inverted_index.NPLM | 90, 127 |
| abstract_inverted_index.also | 136 |
| abstract_inverted_index.test | 77 |
| abstract_inverted_index.that | 125 |
| abstract_inverted_index.when | 33 |
| abstract_inverted_index.with | 27, 105 |
| abstract_inverted_index.Here, | 53 |
| abstract_inverted_index.data, | 37 |
| abstract_inverted_index.data. | 84, 147 |
| abstract_inverted_index.known | 113 |
| abstract_inverted_index.means | 139 |
| abstract_inverted_index.serve | 129 |
| abstract_inverted_index.their | 9, 19 |
| abstract_inverted_index.these | 46 |
| abstract_inverted_index.while | 135 |
| abstract_inverted_index.cases: | 96 |
| abstract_inverted_index.making | 38 |
| abstract_inverted_index.method | 134 |
| abstract_inverted_index.model, | 112 |
| abstract_inverted_index.models | 1, 47, 98, 104 |
| abstract_inverted_index.power, | 30 |
| abstract_inverted_index.proper | 16 |
| abstract_inverted_index.public | 110 |
| abstract_inverted_index.(NPLM), | 64 |
| abstract_inverted_index.Classic | 24 |
| abstract_inverted_index.Machine | 63 |
| abstract_inverted_index.Physics | 61 |
| abstract_inverted_index.applied | 34 |
| abstract_inverted_index.central | 4 |
| abstract_inverted_index.certify | 42 |
| abstract_inverted_index.events. | 122 |
| abstract_inverted_index.modeled | 143 |
| abstract_inverted_index.physics | 120 |
| abstract_inverted_index.propose | 55 |
| abstract_inverted_index.regions | 144 |
| abstract_inverted_index.require | 14 |
| abstract_inverted_index.testing | 70 |
| abstract_inverted_index.through | 21 |
| abstract_inverted_index.trained | 80, 99 |
| abstract_inverted_index.FlowSim, | 115 |
| abstract_inverted_index.Gaussian | 103 |
| abstract_inverted_index.Learning | 62 |
| abstract_inverted_index.approach | 67 |
| abstract_inverted_index.diagnose | 141 |
| abstract_inverted_index.generate | 118 |
| abstract_inverted_index.inspired | 71 |
| abstract_inverted_index.mixtures | 101 |
| abstract_inverted_index.networks | 79 |
| abstract_inverted_index.powerful | 132 |
| abstract_inverted_index.rigorous | 22 |
| abstract_inverted_index.struggle | 26 |
| abstract_inverted_index.benchmark | 95 |
| abstract_inverted_index.collision | 121 |
| abstract_inverted_index.developed | 116 |
| abstract_inverted_index.difficult | 40 |
| abstract_inverted_index.providing | 137 |
| abstract_inverted_index.settings. | 52 |
| abstract_inverted_index.Generative | 0 |
| abstract_inverted_index.approaches | 25 |
| abstract_inverted_index.end-to-end | 111 |
| abstract_inverted_index.generative | 78, 97 |
| abstract_inverted_index.increasing | 106 |
| abstract_inverted_index.realistic, | 49 |
| abstract_inverted_index.scientific | 6, 51, 83 |
| abstract_inverted_index.systematic | 10 |
| abstract_inverted_index.validation | 92, 133 |
| abstract_inverted_index.workflows, | 7 |
| abstract_inverted_index.demonstrate | 86, 124 |
| abstract_inverted_index.high-energy | 119 |
| abstract_inverted_index.limitations | 20 |
| abstract_inverted_index.performance | 88 |
| abstract_inverted_index.reliability | 44 |
| abstract_inverted_index.statistical | 29 |
| abstract_inverted_index.validation. | 23 |
| abstract_inverted_index.increasingly | 3 |
| abstract_inverted_index.scalability, | 28 |
| abstract_inverted_index.construction, | 75 |
| abstract_inverted_index.sub-optimally | 142 |
| abstract_inverted_index.understanding | 17 |
| abstract_inverted_index.interpretation | 13 |
| abstract_inverted_index.learning-based | 66 |
| abstract_inverted_index.Neyman--Pearson | 74 |
| abstract_inverted_index.dimensionality, | 107 |
| abstract_inverted_index.goodness-of-fit | 69 |
| abstract_inverted_index.high-dimensional | 36, 50, 82 |
| abstract_inverted_index.interpretability | 32 |
| cited_by_percentile_year | |
| countries_distinct_count | 0 |
| institutions_distinct_count | 3 |
| citation_normalized_percentile |