{"id":"762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d","title":"arXiv cs.AI: UniGADD, a detector designed to generalize across deepfake generators","argument":"In their paper on arXiv, Bernard Tomczyk, Agil Aghasanli and Plamen Angelov introduce UniGADD, a deepfake detector they say performs well on known generators and more robustly on previously unseen ones. Their method uses two stages of supervised learning. First, a contrastive loss trains the model to distinguish real images from fake images made by known generators. Then it refines the embedding space using real samples, aiming to separate clusters and make samples within each cluster more compact. The authors say this second stage improves generalisability to unseen deepfake techniques. They report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases. The abstract does not provide the datasets, comparison details or measured results, so the extent and scope of that advantage cannot be assessed from this evidence. Escape Meridian interpretation: the proposed detector could help identify fakes from generators not represented in training if its reported results carry over beyond the paper’s evaluation; the abstract does not establish that. Its practical value therefore remains tied to how it performs under the tested conditions and on real-world content.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the arXiv abstract; it does not include datasets, comparison details, measured results or full evaluation conditions. The claim of practical applicability is made by the authors; the abstract does not establish performance on real-world content.","speaker":{"name":"Tomczyk, Bernard, Aghasanli, Agil, Angelov, Plamen","roleAtSource":"Article author"},"source":{"title":"UniGADD: Universal Generator-Agnostic Deepfake Detector","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.07647","publishedAt":"2026-11-05","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce UniGADD, a detector intended to perform on known and previously unseen deepfake generators.","excerpt":"we introduce Universal Generator-Agnostic Deepfake Detector (UniGADD), a novel method that employs supervised learning to achieve high accuracy on known deepfake generators, while maintaining robust performance on previously unseen ones.","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"UniGADD uses two-stage optimisation: contrastive loss learns embeddings from real and known fake images, then a refinement stage operates exclusively on real samples.","excerpt":"In the first stage, a contrastive loss encourages the model to learn discriminative feature embeddings from real and known fake images","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"The authors say the second stage improves generalisability to unseen deepfake generation techniques.","excerpt":"This refinement enhances generalisability, enabling the method to exhibit improved robustness against unseen deepfake generation techniques.","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"The authors report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases.","excerpt":"UniGADD achieves accuracy on par with state-of-the-art methods for known generators, while significantly outperforming them on unseen cases","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"Escape Meridian interpretation: if the reported results carry over beyond the paper’s evaluation, UniGADD could help identify fakes made by generators not represented in training.","excerpt":"while maintaining robust performance on previously unseen ones.","url":"https://arxiv.org/abs/2611.07647","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07647","checkedAt":"2026-11-09T07:50:07.807Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Which evaluation settings and results support the authors’ claim that UniGADD performs significantly better than other methods on previously unseen generators?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: UniGADD, a detector designed to generalize across deepfake generators","summary":"In their paper on arXiv, Bernard Tomczyk, Agil Aghasanli and Plamen Angelov introduce UniGADD, a deepfake detector they say performs well on known generators and more robustly on previously unseen ones. Their method uses two stages of supervised learning. First, a contrastive loss trains the model to distinguish real images from fake images made by known generators. Then it refines the embedding space using real samples, aiming to separate clusters and make samples within each cluster more compact. The authors say this second stage improves generalisability to unseen deepfake techniques. They report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases. The abstract does not provide the datasets, comparison details or measured results, so the extent and scope of that advantage cannot be assessed from this evidence. Escape Meridian interpretation: the proposed detector could help identify fakes from generators not represented in training if its reported results carry over beyond the paper’s evaluation; the abstract does not establish that. Its practical value therefore remains tied to how it performs under the tested conditions and on real-world content.","question":"Which evaluation settings and results support the authors’ claim that UniGADD performs significantly better than other methods on previously unseen generators?","claims":[{"text":"The authors introduce UniGADD, a detector intended to perform on known and previously unseen deepfake generators.","excerpt":"we introduce Universal Generator-Agnostic Deepfake Detector (UniGADD), a novel method that employs supervised learning to achieve high accuracy on known deepfake generators, while maintaining robust performance on previously unseen ones.","support":"source-stated"},{"text":"UniGADD uses two-stage optimisation: contrastive loss learns embeddings from real and known fake images, then a refinement stage operates exclusively on real samples.","excerpt":"In the first stage, a contrastive loss encourages the model to learn discriminative feature embeddings from real and known fake images","support":"source-stated"},{"text":"The authors say the second stage improves generalisability to unseen deepfake generation techniques.","excerpt":"This refinement enhances generalisability, enabling the method to exhibit improved robustness against unseen deepfake generation techniques.","support":"source-stated"},{"text":"The authors report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases.","excerpt":"UniGADD achieves accuracy on par with state-of-the-art methods for known generators, while significantly outperforming them on unseen cases","support":"source-stated"},{"text":"Escape Meridian interpretation: if the reported results carry over beyond the paper’s evaluation, UniGADD could help identify fakes made by generators not represented in training.","excerpt":"while maintaining robust performance on previously unseen ones.","support":"interpretation"}],"limitations":["The supplied evidence is the arXiv abstract; it does not include datasets, comparison details, measured results or full evaluation conditions.","The claim of practical applicability is made by the authors; the abstract does not establish performance on real-world content."],"mediaId":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","correctionReason":null,"id":"762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d","publicationId":"d00cc71c-05b2-4ba0-8d91-2dc2fc99d738","version":1,"sourceUrl":"https://arxiv.org/abs/2611.07647","sourceTitle":"UniGADD: Universal Generator-Agnostic Deepfake Detector","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-05","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-06T05:01:08.464Z","createdAt":"2026-11-08T23:18:08.494Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","kind":"article","url":"https://arxiv.org/abs/2611.07647","sourceUrl":"https://arxiv.org/abs/2611.07647","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-08T23:17:11.320Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07647","checkedAt":"2026-11-09T07:50:07.807Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","confidence":0.74,"provider":"typesafe","model":"jev-1.13.0","inputHash":"8bf29ac665aef681ce7b9cba61a2f8647e03f187362c670a55d9c1e7df54af3b","decidedAt":"2026-11-08T23:17:12.456Z"},"provenance":{"textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observationId":"26e215d0534d17dfa2f87978aa67995d1d5cc5b29ad0f00423b5cbbc85859050","retrievedAt":"2026-11-08T23:17:11.139Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"e20b28f9-2cc1-44d1-8bcc-4c3887207937"},"reviewer":{"model":"gpt-6-luna","runId":"a120114d-4064-412b-837d-3ba599e2f192"}},"audit":{"status":"passed","reviewedAt":"2026-11-08T23:18:08.494Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Tomczyk, Bernard","url":null,"kind":"unknown"},{"name":"Aghasanli, Agil","url":null,"kind":"unknown"},{"name":"Angelov, Plamen","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Tomczyk, Bernard","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Tomczyk, Bernard","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}},{"actor":{"name":"Aghasanli, Agil","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Aghasanli, Agil","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}},{"actor":{"name":"Angelov, Plamen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Angelov, Plamen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}}],"participationCoverage":{"total":3,"returned":3,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.07647","snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z","basis":"publisher-page-metadata","sourceLinks":42,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d/source"}},"canonicalUrl":"https://escapemeridian.com/readings/arxiv-cs-ai-unigadd-a-detector-designed-to-generalize-across-762d6d6b","formats":{"html":"/readings/arxiv-cs-ai-unigadd-a-detector-designed-to-generalize-across-762d6d6b","markdown":"/readings/arxiv-cs-ai-unigadd-a-detector-designed-to-generalize-across-762d6d6b.md","json":"/readings/arxiv-cs-ai-unigadd-a-detector-designed-to-generalize-across-762d6d6b.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d","revision":"d00cc71c-05b2-4ba0-8d91-2dc2fc99d738","revisionKind":"publication-id","title":"arXiv cs.AI: UniGADD, a detector designed to generalize across deepfake generators","url":"https://escapemeridian.com/readings/arxiv-cs-ai-unigadd-a-detector-designed-to-generalize-across-762d6d6b"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.07647","revision":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","revisionKind":"sha256-source-text","title":"UniGADD: Universal Generator-Agnostic Deepfake Detector","url":"https://arxiv.org/abs/2611.07647"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d","title":"arXiv cs.AI: UniGADD, a detector designed to generalize across deepfake generators","argument":"In their paper on arXiv, Bernard Tomczyk, Agil Aghasanli and Plamen Angelov introduce UniGADD, a deepfake detector they say performs well on known generators and more robustly on previously unseen ones. Their method uses two stages of supervised learning. First, a contrastive loss trains the model to distinguish real images from fake images made by known generators. Then it refines the embedding space using real samples, aiming to separate clusters and make samples within each cluster more compact. The authors say this second stage improves generalisability to unseen deepfake techniques. They report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases. The abstract does not provide the datasets, comparison details or measured results, so the extent and scope of that advantage cannot be assessed from this evidence. Escape Meridian interpretation: the proposed detector could help identify fakes from generators not represented in training if its reported results carry over beyond the paper’s evaluation; the abstract does not establish that. Its practical value therefore remains tied to how it performs under the tested conditions and on real-world content.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the arXiv abstract; it does not include datasets, comparison details, measured results or full evaluation conditions. The claim of practical applicability is made by the authors; the abstract does not establish performance on real-world content.","speaker":{"name":"Tomczyk, Bernard, Aghasanli, Agil, Angelov, Plamen","roleAtSource":"Article author"},"source":{"title":"UniGADD: Universal Generator-Agnostic Deepfake Detector","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.07647","publishedAt":"2026-11-05","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce UniGADD, a detector intended to perform on known and previously unseen deepfake generators.","excerpt":"we introduce Universal Generator-Agnostic Deepfake Detector (UniGADD), a novel method that employs supervised learning to achieve high accuracy on known deepfake generators, while maintaining robust performance on previously unseen ones.","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"UniGADD uses two-stage optimisation: contrastive loss learns embeddings from real and known fake images, then a refinement stage operates exclusively on real samples.","excerpt":"In the first stage, a contrastive loss encourages the model to learn discriminative feature embeddings from real and known fake images","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"The authors say the second stage improves generalisability to unseen deepfake generation techniques.","excerpt":"This refinement enhances generalisability, enabling the method to exhibit improved robustness against unseen deepfake generation techniques.","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"The authors report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases.","excerpt":"UniGADD achieves accuracy on par with state-of-the-art methods for known generators, while significantly outperforming them on unseen cases","url":"https://arxiv.org/abs/2611.07647","relationship":"Source-stated"},{"text":"Escape Meridian interpretation: if the reported results carry over beyond the paper’s evaluation, UniGADD could help identify fakes made by generators not represented in training.","excerpt":"while maintaining robust performance on previously unseen ones.","url":"https://arxiv.org/abs/2611.07647","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07647","checkedAt":"2026-11-09T07:50:07.807Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"Which evaluation settings and results support the authors’ claim that UniGADD performs significantly better than other methods on previously unseen generators?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: UniGADD, a detector designed to generalize across deepfake generators","summary":"In their paper on arXiv, Bernard Tomczyk, Agil Aghasanli and Plamen Angelov introduce UniGADD, a deepfake detector they say performs well on known generators and more robustly on previously unseen ones. Their method uses two stages of supervised learning. First, a contrastive loss trains the model to distinguish real images from fake images made by known generators. Then it refines the embedding space using real samples, aiming to separate clusters and make samples within each cluster more compact. The authors say this second stage improves generalisability to unseen deepfake techniques. They report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases. The abstract does not provide the datasets, comparison details or measured results, so the extent and scope of that advantage cannot be assessed from this evidence. Escape Meridian interpretation: the proposed detector could help identify fakes from generators not represented in training if its reported results carry over beyond the paper’s evaluation; the abstract does not establish that. Its practical value therefore remains tied to how it performs under the tested conditions and on real-world content.","question":"Which evaluation settings and results support the authors’ claim that UniGADD performs significantly better than other methods on previously unseen generators?","claims":[{"text":"The authors introduce UniGADD, a detector intended to perform on known and previously unseen deepfake generators.","excerpt":"we introduce Universal Generator-Agnostic Deepfake Detector (UniGADD), a novel method that employs supervised learning to achieve high accuracy on known deepfake generators, while maintaining robust performance on previously unseen ones.","support":"source-stated"},{"text":"UniGADD uses two-stage optimisation: contrastive loss learns embeddings from real and known fake images, then a refinement stage operates exclusively on real samples.","excerpt":"In the first stage, a contrastive loss encourages the model to learn discriminative feature embeddings from real and known fake images","support":"source-stated"},{"text":"The authors say the second stage improves generalisability to unseen deepfake generation techniques.","excerpt":"This refinement enhances generalisability, enabling the method to exhibit improved robustness against unseen deepfake generation techniques.","support":"source-stated"},{"text":"The authors report accuracy on par with state-of-the-art methods for known generators and significantly better performance on unseen cases.","excerpt":"UniGADD achieves accuracy on par with state-of-the-art methods for known generators, while significantly outperforming them on unseen cases","support":"source-stated"},{"text":"Escape Meridian interpretation: if the reported results carry over beyond the paper’s evaluation, UniGADD could help identify fakes made by generators not represented in training.","excerpt":"while maintaining robust performance on previously unseen ones.","support":"interpretation"}],"limitations":["The supplied evidence is the arXiv abstract; it does not include datasets, comparison details, measured results or full evaluation conditions.","The claim of practical applicability is made by the authors; the abstract does not establish performance on real-world content."],"mediaId":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","correctionReason":null,"id":"762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d","publicationId":"d00cc71c-05b2-4ba0-8d91-2dc2fc99d738","version":1,"sourceUrl":"https://arxiv.org/abs/2611.07647","sourceTitle":"UniGADD: Universal Generator-Agnostic Deepfake Detector","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-05","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-06T05:01:08.464Z","createdAt":"2026-11-08T23:18:08.494Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","kind":"article","url":"https://arxiv.org/abs/2611.07647","sourceUrl":"https://arxiv.org/abs/2611.07647","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-08T23:17:11.320Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07647","checkedAt":"2026-11-09T07:50:07.807Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"b8549baa8cbec9408cffe85644f73a9cc591c950864c05591b886a9cd52539e1","confidence":0.74,"provider":"typesafe","model":"jev-1.13.0","inputHash":"8bf29ac665aef681ce7b9cba61a2f8647e03f187362c670a55d9c1e7df54af3b","decidedAt":"2026-11-08T23:17:12.456Z"},"provenance":{"textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observationId":"26e215d0534d17dfa2f87978aa67995d1d5cc5b29ad0f00423b5cbbc85859050","retrievedAt":"2026-11-08T23:17:11.139Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"e20b28f9-2cc1-44d1-8bcc-4c3887207937"},"reviewer":{"model":"gpt-6-luna","runId":"a120114d-4064-412b-837d-3ba599e2f192"}},"audit":{"status":"passed","reviewedAt":"2026-11-08T23:18:08.494Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Tomczyk, Bernard","url":null,"kind":"unknown"},{"name":"Aghasanli, Agil","url":null,"kind":"unknown"},{"name":"Angelov, Plamen","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Tomczyk, Bernard","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Tomczyk, Bernard","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}},{"actor":{"name":"Aghasanli, Agil","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Aghasanli, Agil","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}},{"actor":{"name":"Angelov, Plamen","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Angelov, Plamen","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07647","locator":"head/meta[name=author|citation_author]"},"assertionId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T07:50:07.807Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","sourceUrl":"https://arxiv.org/abs/2611.07647","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z"}}],"participationCoverage":{"total":3,"returned":3,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.07647","snapshotId":"921855925157029d6045620a476b53f4c59baa8764e53b83e7a35f4b9fd29c03","contentSha256":"412b0715a89caefd1c845c497b16e8806bb6937b3c302d7d64b80aec8be93a8b","textHash":"05e109f221ad2a14dec200d7d5a85b00cad58be4bae4fb60d8a9939a885d9d52","observedAt":"2026-11-09T07:50:07.807Z","basis":"publisher-page-metadata","sourceLinks":42,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/762d6d6b447cc51106ee4575f0588c962032b04df104285de747e24d029c910d/source"}}}],"audits":[{"publicationId":"d00cc71c-05b2-4ba0-8d91-2dc2fc99d738","verdict":"passed","findings":[{"dimension":"attribution","status":"pass","explanation":"The paper is correctly attributed to Bernard Tomczyk, Agil Aghasanli and Plamen Angelov on arXiv. Performance claims remain framed as the authors' reports. The title's arXiv cs.AI label matches the publisher field. The paper's primary category is cs.CV, cross-listed in cs.AI.","excerpt":"Authors:Bernard Tomczyk, Agil Aghasanli, Plamen Angelov"},{"dimension":"summary","status":"pass","explanation":"The summary leads with the actor and the development: a detector meant to work on deepfake generators not seen in training. It then describes the two-stage method and the reported results, and states what the abstract leaves unknown. Terms such as contrastive loss and embedding space are kept exactly; tooltips could explain them. The central idea is still understandable without them. The question is a genuine open issue that follows from the missing evaluation details.","excerpt":"UniGADD achieves accuracy on par with state-of-the-art methods for known generators, while significantly outperforming them on unseen cases"},{"dimension":"date","status":"pass","explanation":"The originalPublished date of 2026-11-05 matches the v1 submission date on arXiv. The later feed entry on 2026-11-06 is correctly not used.","excerpt":"[Submitted on 5 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"Each claim is backed by the abstract. The second stage is correctly described as applied only to real samples. The Escape Meridian interpretation is explicitly conditional, labeled, and kept separate from the evidence. The limitations accurately note that the abstract gives no datasets or measured results. Calling the method two stages of supervised learning is a close paraphrase of supervised learning plus two-stage optimisation, not a material distortion.","excerpt":"In the second stage, the embedding space is refined by promoting inter-cluster separation and intra-cluster compactness, applied exclusively to real samples."},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the original arXiv abstract page. This is the most direct access to the source.","excerpt":null}],"createdAt":"2026-11-08T23:18:08.494Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}