{"as_of":"2026-08-07T21:22:00Z","caps":{"database_statements":6,"inbound":100,"outbound":100},"context_digest":"sha256:03501f9bd520f7a9a2afa07a0648fb2f1205a514572bca7e42ddf3a0874d0390","coverage":[{"denominator":89,"lane":"reference_resolution","note":"Typed states for the displayed outbound observations.","records_observed":89,"source":"paper_references, paper_reference_links","source_observed_at":"2026-07-12T08:12:18.373103Z","state":"measured"},{"denominator":89,"lane":"standing_notices","note":"One-hop event checks from named stored sources.","records_observed":89,"source":"scholarly_work_events, retraction_status_cache","source_observed_at":"2026-08-07T06:34:17.273281+00:00","state":"measured"},{"denominator":0,"lane":"inbound_itemization","note":"Pith citing papers itemized under the disclosed page cap.","records_observed":0,"source":"paper_references, paper_reference_links","source_observed_at":null,"state":"measured"},{"denominator":1,"lane":"external_citation_measurements","note":"A source-named dated measurement, never combined with another source.","records_observed":0,"source":"cited_works","source_observed_at":null,"state":"measured"}],"external_citation_measurements":[],"inbound":[],"links":{"evidence":"/evidence","html":"/paper/2607.02637/citation-record","integrity":"/paper/2607.02637/integrity","json":"/paper/2607.02637/citation-record.json","paper":"/paper/2607.02637"},"outbound":[{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Toward understanding generative data augmentation","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":1,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:aef875325752d0c26dbbc42b881a8972c0f6a8b7a7c1cea0ff9960bb8dab27ba","observation_id":"e1355a76-1825-4b92-800c-a24a321c0c1e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Fake it till you make it: Learning transferable representations from synthetic imagenet clones","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":2,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:9300a11fbe179b6b17735ec585893c048df7184adc90a8627aa3df1993d31f63","observation_id":"b0040ee0-1618-4b32-aef5-256a16bdfbc0","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":3,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:1b46cd09438a174fca6bba4106d753d8186eb86c472a46f073834957adfb2cab","observation_id":"d4adee76-9cc6-4d31-967f-8da41b43cf27","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Is synthetic data from generative models ready for image recognition? In The Eleventh International Conference on Learning Representations, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":4,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:fcd2afe2fefb32f6fa3026a56a38d73e49dc86171a0c5378c262271174459a31","observation_id":"75f1fa29-079b-496b-9735-bee8fccbe2e1","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"IS SYNTHETIC DATA USEFUL FOR TRANSFER LEARNING? AN INVESTI- GATION INTO DATA GENERATION, VOLUME, AND UTILIZATION, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":5,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:7288cdd96f839b9acea5e42e76d56fa0154e78b8602d050f83eab89747a17abb","observation_id":"0d2c5211-a986-49b6-ace6-cecaa2330971","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Scaling laws of synthetic images for model training","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":6,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0fa54a9bc23ba1720c26994d35fece566d65cfdc689b9e1ff6ee59c255d77db0","observation_id":"023d64d6-e998-4e95-bccf-3bd68770886f","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Datadream: Few-shot guided dataset generation","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":7,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:d1453be5186e938a23abbcd903e935fe68faceb1fa8d123ceda3f4c155653e11","observation_id":"7fc2659c-c68a-4188-9145-e1a3c13844d4","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Real-fake: Effective training data synthesis through distribution matching, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":8,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0921ddf5f0f7cf2b388b63da6520a3fecd5e66dc180963a620f3dc11ba4b69e3","observation_id":"c971d12c-d49b-4963-ae13-54842c6cff96","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.05803","last_updated":"2025-04-17T12:23:46Z","snapshot_observed_at":"2026-07-06T20:19:07.706492Z","submitted_at":"2025-01-10T09:10:30Z","title":"Test-time Alignment of Diffusion Models without Reward Over-optimization","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.05803","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Test-time alignment of diffusion models without reward over-optimization.arXiv preprint arXiv:2501.05803, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":9,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2501.05803","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6c853a92c44a15cafbf7dd9047255124e4f9f4b91d38cf5f0ef7ae92f6546b87","observation_id":"a43922f5-fe81-43aa-ac45-82319f87d8a4","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Diffusion curriculum: Synthetic-to-real data curriculum via image-guided diffusion","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":10,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6598dee747b7f840dcf08863abefdb954b2e4ed02305cc70a8e0ed05dd64e5db","observation_id":"06efc3cc-0267-4901-b6ac-ac0bc43f367e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Increasing the utility of synthetic images through chamfer guidance.arXiv preprint arXiv:2508.10631, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":11,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0fc8e04f838907b5cdf51429688471b3bfeb9012ecc17f6a1e038c926f97e6d3","observation_id":"3f74dd41-0a71-4898-8edb-bcea2b497560","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2410.06025","last_updated":"2025-05-28T17:28:20Z","snapshot_observed_at":"2026-08-07T11:15:36.426707Z","submitted_at":"2024-10-08T13:26:32Z","title":"Shielded Diffusion: Generating Novel and Diverse Images using Sparse Repellency","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2410.06025","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Shielded diffusion: Generating novel and diverse images using sparse repellency.arXiv preprint arXiv:2410.06025, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":12,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2410.06025","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:09b0d733020b83b83f14f82d97d5c51be25fb6eda2f2b7cd25315682d815a465","observation_id":"da328825-6f83-4bc5-bf34-7967acb23b31","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Diversity is definitely needed: Improving model-agnostic zero-shot classification via stable diffusion","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":13,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:f4b51a7201041e7b586a30eda8f330cffe29a96b75407420c61d576e7335a683","observation_id":"09e81351-233f-4430-b173-7485236a1047","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Data augmentation for image classification using generative ai","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":14,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:c968a04546b2ff9886c7c797d6082608cdf76fe5340aa028597c30cd015091dc","observation_id":"b68d655a-4db1-417c-b6b5-bc9d5a5eccd5","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Gonzalez, and Trevor Darrell","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":15,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:33f0b1dfba59c1852b6493cbb25c8aca21362f2e7cb16c1e812d7c1954233087","observation_id":"4b8b0bfc-671c-4503-8fe2-8934afddcbba","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Datacomp: In search of the next generation of multimodal datasets","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":16,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6faa8bb66df8fbd5af395213a2a32fc58461eb0e74c8c51c4cff0ac935a47f15","observation_id":"16792d31-3f06-4c0b-bc1d-0ca6cab6ac43","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2303.09540","last_updated":"2023-03-22T17:22:35Z","snapshot_observed_at":"2026-08-06T15:07:40.203199Z","submitted_at":"2023-03-16T17:53:24Z","title":"SemDeDup: Data-efficient learning at web-scale through semantic deduplication","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2303.09540","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Semd- edup: Data-efficient learning at web-scale through semantic deduplication.arXiv preprint arXiv:2303.09540, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":17,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2303.09540","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0d3ecfa01f59746c1088333011b4bab3d4852caedc180d111ddb83722469cb21","observation_id":"bd0cdebc-125f-4f48-9fbe-23c03c040cf9","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Effective audio classification network based on paired inverse pyramid structure and dense mlp block","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":18,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:b2c6aa77c79a867817a450c9130c94877bde5453f47a6856330e06bdb363f839","observation_id":"06ff4a81-3f80-4af4-bbf7-d9e8bd2bfd74","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Datasetgan: Efficient labeled data factory with minimal human effort","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":19,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:66eebb73f3e1c417c2d04325f95674d2ebef7443afb53322acbeb69f5649b749","observation_id":"fac906f7-8c75-4b0f-8e54-81e8370727ac","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Improved precision and recall metric for assessing generative models","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":20,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:47d6fdf02906a0a5898fc437e4f775be5c365694027395214d1878078c1fa6ab","observation_id":"be454cb5-4a19-43a4-a95e-b86a4b77227c","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Explore the power of synthetic data on few-shot object detection","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":21,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:ac906edb42e9280df2caec8dc63cebe2d24fe9d0302a226190ae7ebb5604e3f1","observation_id":"20d55aa3-9361-4426-ab66-b4a554f74074","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Aerogen: Enhancing remote sensing object detection with diffusion-driven data generation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":22,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:46b9bdd508bd7fcf821f273170ddd6f4222c8c8efbb79c733fb6a5dc621103a3","observation_id":"e53703b0-8ccc-4ccd-9e54-6b75f046c869","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Noise-consistent siamese-diffusion for medical image synthesis and segmentation","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":23,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6b8396acf2f7490f4d146c4a463480d2d5c115156766455d8d1fc6224a0262e3","observation_id":"701fbeee-3d29-4f18-8fb7-993c7adf60fb","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Domain gap embeddings for generative dataset augmentation","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":24,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6be725be91b3971c41fd345573d3ab55acac04d6cf8a119a63ab5369b9e74b8f","observation_id":"cb39eda3-b953-4f45-a4ba-f35afe65198f","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2305.15316","last_updated":"2023-05-24T16:33:02Z","snapshot_observed_at":"2026-08-05T03:25:04.857257Z","submitted_at":"2023-05-24T16:33:02Z","title":"Training on Thin Air: Improve Image Classification with Generated Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2305.15316","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Training on thin air: Improve image classifica- tion with generated data, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":25,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2305.15316","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:dd8c4b6216c0a9b5e7180b19d361f8bf0488feae859f473b0f178fbb4d166bad","observation_id":"64a7a392-e6e5-4735-a384-1e2a5a38a417","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2306.00984","last_updated":"2023-10-26T15:16:57Z","snapshot_observed_at":"2026-07-06T15:36:45.014921Z","submitted_at":"2023-06-01T17:59:51Z","title":"StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2306.00984","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Stablerep: Synthetic images from text-to-image models make strong visual representation learners, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":26,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2306.00984","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:7ecd19e154f3b9bb8662e6bddb3b6bd8a5d013a31041ea13a24e908a43311358","observation_id":"3d459ab5-3ac5-41ac-8076-8c320fc6801b","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2402.01832","last_updated":"2024-07-18T10:21:29Z","snapshot_observed_at":"2026-08-04T15:45:57.991866Z","submitted_at":"2024-02-02T18:59:58Z","title":"SynthCLIP: Are We Ready for a Fully Synthetic CLIP Training?","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2402.01832","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Synthclip: Are we ready for a fully synthetic clip training?, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":27,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2402.01832","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0d88131d009d4fe3929cde008d0ad4747603b9d15c4d51527c2b957c1bd25f07","observation_id":"aa85cf8d-ae67-46d6-a24b-86508a331053","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.17742","last_updated":"2023-12-28T18:59:55Z","snapshot_observed_at":"2026-07-06T17:09:59.848387Z","submitted_at":"2023-12-28T18:59:55Z","title":"Learning Vision from Models Rivals Learning Vision from Data","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.17742","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Learning vision from models rivals learning vision from data, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":28,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2312.17742","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:96b954afa89684b5798f2de75c5367fee3bfdf3246f6e888723283d68366c94f","observation_id":"3eb9119a-aa39-4608-ad89-b239cf7e0b55","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2408.16965","last_updated":"2025-04-24T16:00:27Z","snapshot_observed_at":"2026-08-05T22:08:30.450143Z","submitted_at":"2024-08-30T01:47:43Z","title":"Contrastive Learning with Synthetic Positives","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2408.16965","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Contrastive learning with synthetic positives, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":29,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2408.16965","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:c4a5e5e4ab5a042428586d9f6e2b8b4465206b4ee96d279c0eea246ab785c040","observation_id":"5c857f26-fb2b-4a17-9f99-fa7c40d431a4","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2308.04553","last_updated":"2024-07-17T15:07:01Z","snapshot_observed_at":"2026-08-05T19:01:22.053485Z","submitted_at":"2023-08-08T19:52:28Z","title":"From Fake to Real: Pretraining on Balanced Synthetic Images to Prevent Spurious Correlations in Image Recognition","version":3},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2308.04553","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":30,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2308.04553","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:a342991e46e99f797802b37cae0bf9612cee1e23c785f891dd8c3a1ec5777005","observation_id":"a4edce16-eeec-482d-9c97-a178e98647a0","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Will large-scale generative models corrupt future datasets? In2023 IEEE/CVF International Conference on Computer Vision (ICCV), page 20498–20508","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":31,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:e861caf2d890c60c5374ea189df3397cabc5b62af946a713373c78d4c32ba3fc","observation_id":"2f3a24af-b873-4412-a4bf-c0b2491d4ad1","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Do generated data always help contrastive learning?,","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":32,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:42da342786505c8361c1ed48b8580f4878b3f040470bc7882b7b81ffb6db0237","observation_id":"88f393b9-784d-42ab-9a20-510e9ecd2f60","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2403.12448","last_updated":"2024-03-19T05:17:47Z","snapshot_observed_at":"2026-07-06T17:46:50.682231Z","submitted_at":"2024-03-19T05:17:47Z","title":"Do Generated Data Always Help Contrastive Learning?","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2403.12448","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":33,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2403.12448","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:162bae1b4fd14cfdfbb76d99778c8fdac0beea39eb05867bcd0cf1feab8def86","observation_id":"40335866-76d5-4e75-a5ad-f1c1cf3dd289","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Utilgen: Utility-centric generative data augmentation with dual-level task adaptation","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":34,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:2b9ab4ffcc4af09fe94889cf5f1637d4af756ab1470ba259dcdd6ec1e0e9da0f","observation_id":"3ec97765-2d95-4e05-a482-e97099beef35","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":35,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:525045c7755ef584090ad6ecca98db834401edee88e5893a0786344b3a73ad07","observation_id":"cb8dd1a1-5163-4608-89d4-0c095d1684b9","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2401.04578","last_updated":"2024-03-12T10:35:56Z","snapshot_observed_at":"2026-07-06T17:13:17.263599Z","submitted_at":"2024-01-09T14:32:24Z","title":"Effective pruning of web-scale datasets based on complexity of concept clusters","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2401.04578","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Effective pruning of web-scale datasets based on complexity of concept clusters","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":36,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2401.04578","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:a271106fc2a9357a2b84061bd9e7d5d6213c5363374c7e8d80d3dac1025c9ffc","observation_id":"5102bbca-3efa-42be-ad98-6c3fcb2d8602","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Learning what matters: Prioritized concept learning via relative error-driven sample selection.arXiv preprint arXiv:2506.01085, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":37,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6bf93189bade4f2bd10a6cc92fa9d416621afc3738506265e3350da475c21401","observation_id":"7ddd59c0-5359-425d-9ff3-743e332b4780","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Sampling strategies for gan synthetic data","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":38,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:4c243ac9610e7f139b9ecb4810d294841e01f122671e6b8cc1b9c6f0a1b5ea16","observation_id":"1f31d0fb-7ed7-44b5-ad15-37ad28aa88b5","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Datasetgan: Efficient labeled data factory with minimal human effort","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":39,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:cebc61d56738195f14452284593b2ff35fba4e9420cc3c195cf7d36b38fbb590","observation_id":"bc68cf69-693c-4273-96e6-378cdf4c7b75","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Data aug- mentation for environmental sound classification using diffusion probabilistic model with top-k selection discriminator","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":40,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:52c21d269b8cfd5618617ae291b7bbab9ff6afac85205fdf515aafae476cfe16","observation_id":"bbc878a9-434e-4e76-a39b-50fab4d3db5c","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Strata: Self-training with task augmentation for better few-shot learning","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":41,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:07b32eca750f94248a0dab165a26fdd582f7a2ecd6cbd042599b08308ebe1e4c","observation_id":"0a36b55b-4443-4752-9533-9e799cd9b562","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":"10.18653/v1/2021.emnlp-main.462","metadata_source":"openalex","pith_arxiv_id":null,"snapshot_observed_at":"2026-08-05T02:28:24.338817Z","title":"URL http://dx.doi.org/10.18653/v1/ 2021.emnlp-main.462","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","work_id":"611e5503-17ab-4683-8e05-9c4197c5596b","year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":42,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:82c5782b5dc44c785910a8511b04b2ddf78addc2754aeeef0b0cb32d5985b5f2","observation_id":"74761493-109e-4da7-bbba-f603e8762252","resolution":{"observed_at":"2026-07-12T08:18:38.055039Z","resolver_source":"doi","status":"verified_exact"},"standing_notice":{"events":[],"observation":"No event found in the named queried sources as of 2026-08-07T06:34:17.273281+00:00.","reason":null,"source_receipts":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"state":"measured"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Learning transferable visual models from natural language supervision","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":43,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:d5ed5526a21cf148ada6d1eeced947360d64e19de7f4cd1fe4b4f29d7ecc5cb2","observation_id":"75a5b133-6f3f-4530-8fd5-34acb5d86302","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2501.15201","last_updated":"2025-01-25T12:53:14Z","snapshot_observed_at":"2026-08-06T19:09:08.132113Z","submitted_at":"2025-01-25T12:53:14Z","title":"A Training-free Synthetic Data Selection Method for Semantic Segmentation","version":1},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2501.15201","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"A training-free synthetic data selection method for semantic segmentation, 2025","venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":44,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2501.15201","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:e076548347448ca956fa3039e894a426e294980ec967c5bfd91fc2b05b2ed410","observation_id":"3c66ea9f-3186-4c3f-a7c0-248261da1b5c","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Reliable fidelity and diversity metrics for generative models, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":45,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:0020423871d74db454e32aba32f8628ba221b92ee673c53a7ab0118ce4a7e332","observation_id":"0f691097-9cf5-48ed-a8a8-2e2bcc1f2622","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Deep data augmentation for weed recognition enhancement: A diffusion probabilistic model and transfer learning based approach","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":46,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:71d66eb763f0f6d63f08a293288011bf9969e697241cf433f8fe3573c1bbbeb5","observation_id":"4b1733a7-16b7-4bc7-8970-980bcf9e2457","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2312.03046","last_updated":"2023-12-07T02:04:49Z","snapshot_observed_at":"2026-08-07T15:26:35.914331Z","submitted_at":"2023-12-05T17:18:09Z","title":"Diversified in-domain synthesis with efficient fine-tuning for few-shot classification","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2312.03046","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Diversified in-domain synthesis with efficient fine-tuning for few-shot classification.arXiv preprint arXiv:2312.03046, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":47,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2312.03046","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:7b204a447219a255f86f0c2381b25c847718531c32f3ab62f90a95c119611693","observation_id":"ab170a54-206a-4c30-ab61-36acf4682d74","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Effective data augmentation with diffusion models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":48,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:327fb66f28cad8e1264ebbdfd847bf772d2f0ddc2fc0fb1b83408c3de07c6205","observation_id":"3138cb68-7f93-4c53-8755-673dee28d62e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2310.00158","last_updated":"2024-09-09T22:06:05Z","snapshot_observed_at":"2026-08-04T13:07:43.258940Z","submitted_at":"2023-09-29T21:47:57Z","title":"Feedback-guided Data Synthesis for Imbalanced Classification","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2310.00158","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Feedback-guided data synthesis for imbalanced classification, 2024","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":49,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2310.00158","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:bd08b321d19c5a2c892c340000d868718ddab6015225fb768be3545f3056c833","observation_id":"27707d05-393f-4eef-a97b-bf0ca7920cd2","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2104.02057","last_updated":"2021-08-16T17:40:21Z","snapshot_observed_at":"2026-07-06T10:56:39.571860Z","submitted_at":"2021-04-05T17:59:40Z","title":"An Empirical Study of Training Self-Supervised Vision Transformers","version":4},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2104.02057","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"An empirical study of training self-supervised vision transformers.arXiv preprint arXiv:2104.02057, 2021","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":50,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2104.02057","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:92e4eeb392b2542d4ac19db095811e58ca4e42a4f9df9f59b3ad6346bd581d65","observation_id":"ba36f626-90bb-4bbf-a811-5e001edf2eaa","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2011},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":51,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:3af9ba9994d519f618bd46c1b58c3e24fee3ee84ef1a63ea1bf1f96a96472b5a","observation_id":"67cdcc85-26b1-4e46-ba97-82d5c266a5b9","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Cifar-10 (canadian institute for advanced research).URL http://www","venue":null,"work_id":null,"year":2010},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":52,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:f0da60b801c69b82b9afa507bba104996a42ca4c3ae4f6649894bd950a360161","observation_id":"a7fd599a-caa0-4a44-a6d1-7e486fb23fc1","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":53,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:28a8a3483a2fd18e50c324d59516d38e34bd27ff45ef6cb7554040169dffd9b0","observation_id":"1645fe86-f844-45da-b47d-20c9f3755c19","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Better diffusion models further improve adversarial training","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":54,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:1338b6e6c22a5dd776c5110f88129f56e8d3dfa98b5ecff57a2ddbe40fb6a668","observation_id":"915afd32-b346-485f-9e23-9587f5e683b2","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Analyzing and improving the training dynamics of diffusion models","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":55,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:4abeebbc5b00502a5c275cca2311057c7944296196f909a439a0b4f32dc5cba9","observation_id":"cdfd6e3f-b029-4178-812f-b60cc2651529","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Guiding a diffusion model with a bad version of itself","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":56,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:cfd6ca82bd00f9596ea0374b4d4dea27de53016511239fb662f87205a62d0672","observation_id":"6b221a5b-f291-4d7d-b59d-86e1316069c1","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Contrastive multiview coding","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":57,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:8ff86e71b50c48f56de2feffad4243832f59dd459b5a9dfdcae0f3b392892c9e","observation_id":"29302c09-3c78-4e7a-9c0a-165fc331a393","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2015},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":58,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:197907debc1f2ece6ab4b6c7f726c1053978b1968aae0b43b5e01382980f030b","observation_id":"6b56dd3f-ff34-4f6c-a536-166edc5e4920","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"2109.15102","last_updated":"2021-10-05T11:44:47Z","snapshot_observed_at":"2026-07-06T11:53:04.443134Z","submitted_at":"2021-09-30T13:07:04Z","title":"Fake It Till You Make It: Face analysis in the wild using synthetic data alone","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"2109.15102","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Cashman, and Jamie Shotton","venue":null,"work_id":null,"year":2021},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":59,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/2109.15102","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:2b7d375cd82a0b16bb26d6e91a6e1775789b34fa8a4cf0474f9ab88778a2b660","observation_id":"e27fa6d7-eb42-47ea-9195-d8360478c004","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Extracting training data from diffusion models","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":60,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:5cd10a261f64c388f83266005a3746176b429b8023fd8d6759d95e1257dcca44","observation_id":"c7c9c9d5-9ed1-45dc-a9a8-b60a1f2b0367","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Dcface: Synthetic face generation with dual condition diffusion model","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":61,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:12fb5b043c6f60b1dea4e2eee44e519bb2e1e9780b85ad3872782dd628b19856","observation_id":"80b11d88-b168-479e-bf72-76cd3005c393","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Improving geo-diversity of generated images with contextualized vendi score guidance","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":62,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:222be7417710f08a51e121c352e0fa8cc6803e7de55780a2c8153a63de2ff2a7","observation_id":"10718b42-68bd-464d-a969-cb99d6a14daa","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Jodiffusion: Jointly diffusing image with pixel-level annotations for semantic segmentation promotion","venue":null,"work_id":null,"year":2026},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":63,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:e065c84f9ad4f466c81c7f862b59ad6b498fc72767fddf091afeffd4a9ceb84c","observation_id":"ad2a931e-c74f-4cbe-ac0f-744e30e54d4b","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Deep residual learning for image recognition","venue":null,"work_id":null,"year":2016},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":64,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:bf095b8d34f2e5540097b8db0f386f86a42de8cd78fe513cfe12984bf805c1ca","observation_id":"fabf96b1-c064-451a-b31c-619a7488f66d","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Efficientnet: Rethinking model scaling for convolutional neural networks","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":65,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:5f5271064c733f4fa855a5ddc377b3681c6376104a76cf9821b1f9e618d86be2","observation_id":"a54c39b0-b0fd-46e8-8f47-44a8930dcd1d","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Visual transformers: Token-based image representation and processing for computer vision, 2020","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":66,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:c9ae2494bf2f749d01a376230d493505dbea024b4173d6ce6f1ac805f7b82572","observation_id":"16c800b7-8b16-4739-9f3b-40c2bc93a47e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Benchmarking neural network robustness to common corruptions and perturbations","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":67,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:f7267bd1fc08cc25330acc6a72689fec2d46da4634f41a62a24634dc67793bbc","observation_id":"ecaaa5b2-8d24-4868-a6d7-9b94e4708068","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":{"arxiv_id":"1902.10811","last_updated":"2019-06-12T17:42:33Z","snapshot_observed_at":"2026-08-07T19:33:21.437256Z","submitted_at":"2019-02-13T20:35:44Z","title":"Do ImageNet Classifiers Generalize to ImageNet?","version":2},"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":"1902.10811","snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Do imagenet classifiers generalize to imagenet?, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":68,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"cited_paper":"/paper/1902.10811","citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:cce1928ea152a03e667deedd2e5269ba84a196d7b1c165ab9fead992b83247e9","observation_id":"c991d3c9-be5b-451c-a204-dbde66e65342","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Learning robust global represen- tations by penalizing local predictive power","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":69,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:8a676e06d39f5ef2dd9e46ff32d4fe10a9a9b9d757f804a7b467448535037002","observation_id":"526058aa-7cee-4b18-9289-cb7246097b13","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Benchmarking neural network robustness to common corruptions and perturbations.Proceedings of the International Conference on Learning Representations, 2019","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":70,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6e86fa9b9a556d6fa96b7a533baef4c269f5b6403bc3c89da1408d5456405509","observation_id":"087933fe-beeb-47d1-b2b8-860c138cb67e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Imagenet-cartoon and imagenet-drawing: two domain shift datasets for imagenet","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":71,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:07a0978a2aef796f0111c465a15fa33a32df426456f582d539a5bb7e2ad71871","observation_id":"5645ab9e-19df-424a-9de5-4d9b31de67c1","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Cifar- 10-warehouse: Broad and more realistic testbeds in model generalization analysis","venue":null,"work_id":null,"year":2024},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":72,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:fbe97740d99d95b471217d24d3d1334833ab2ded6fdb3b6eac83678e7f1d94c9","observation_id":"cb15076a-446a-4a51-ac8a-9e1c08c6809e","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Encoder-decoder with atrous separable convolution for semantic image segmentation","venue":null,"work_id":null,"year":2018},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":73,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:fcef483bbd1eafebe43b325f700f80a5d771a968666b8a3b429aae78eb56a075","observation_id":"f7415725-497e-4e16-8399-a4f5550059cf","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Training generative adversarial networks with limited data","venue":null,"work_id":null,"year":2020},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":74,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:97e3a83534b887a0672a5e991841dbb9ade912cc6ab0c7b9c1eaeab5ab69ced8","observation_id":"22567e67-157c-4098-a6be-1607a1b88357","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Elucidating the design space of diffusion-based generative models","venue":null,"work_id":null,"year":2022},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":75,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:7d39d567a8467bb88e37290bc34e60e165b17cfbff0b3876d16a0584d4411021","observation_id":"f035b458-479d-4c57-be9f-73efa8e5ec7f","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Sigmoid loss for language image pre-training, 2023","venue":null,"work_id":null,"year":2023},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":76,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:404b29cb9954b0deb1821fe874fd710174a18bdddc066af320e61ce820c15fcc","observation_id":"bc4d270b-0e46-4383-9fbc-36611f921309","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":2025},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":77,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:3198d1a96f37b5b7a5d8c40f80adf1c0c4f25fb36fe1a6957ecbc1497b173816","observation_id":"7f83ac2a-5019-482c-a525-30f302e90cba","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"near” set contains canonical patterns, while the “far","venue":null,"work_id":null,"year":2019},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":78,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:5110e30d7b910e302de8aa28a3b80b0623f0743652ae6d5c196740958566c661","observation_id":"26f3c55a-49dc-4604-8e23-c5431f54fe90","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Such property implies that learning the pattern in HO, and then we can reconstruct the whole original feature space with the smallest cost","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":79,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:53060d7eab78820eb08567734576ecd3fc3eac096447b0293cb13a2e8985e7b2","observation_id":"48e1a056-11e1-4925-944e-4e65091e8fc7","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"$NUM_GPU","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":80,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:edc122c807c0bb505d48a2c47b15ac33d14a6048595b573a863d0ba39eecdc6b","observation_id":"9ebe5f9e-7b24-467a-9cc4-b2543bc34c0b","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"malformed_identifier"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":81,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:5bb20ca676999431fa4b572fa0e595b0996441f3b43fbf3d413cb5db8f8dee94","observation_id":"e4a56cfd-ffa0-4989-928f-529ea4413c85","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":82,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:6999efa2c9ae7a3c562cdec1a58cab071b60457b2db7563bcc04ae8ab3a7a48a","observation_id":"7afb7a9a-54cd-4211-81d2-25e710c94a0f","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":83,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:d2a6ca23c20a6d2cfa991f7fc55e384ed2cafe7ccf2e64299ca894d89ad9dddc","observation_id":"8f9329c6-7ae7-47df-88d4-2b0995b73920","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":84,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:66e514b3a26dab2c80f2e1a751132ef54e38de2be1befd728ab8032de5467574","observation_id":"915aa267-292d-4ff1-8c8b-a21db6773702","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"common_image","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":85,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:2b0a881bd13178afb9fa093b62c22aa46915dc400d679fcbfde8f18ab6d4f246","observation_id":"7f5b7eb1-9956-476a-b0e1-040be514d338","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":null,"venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":86,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:3c3e689505c0038e0cd5c556db9b781b8b25fa82b8d9defa5e309e0a400502a4","observation_id":"90c21052-07e9-4c1c-8b76-bb8e95b1225b","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"As vision-language models continue to advance,13 synthetic datasets that support multimodal training are becoming increasingly important","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":87,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:78b741238623e7777124e509edfa8d68d72c9552e9da21b5e0de1ed33675a658","observation_id":"34ea2102-4edc-46bf-9836-7eca6caaaade","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"In practice, however, only a small19 reference set is often sufficient, since the partition depends more on relative intra-class similarity than20 on absolute data scale.21","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":88,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:687073eba0f925de56fdde2b467c299e31fcf9ccc0026c1af15ec00f48e7e14d","observation_id":"38464103-6ed1-4cbc-86b0-ccfa133d65c4","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}},{"citation":{"cited_paper":null,"cited_work":{"arxiv_id":null,"doi":null,"metadata_source":null,"pith_arxiv_id":null,"snapshot_observed_at":"2026-07-12T08:12:18.373103Z","title":"Although Fig","venue":null,"work_id":null,"year":null},"citing_paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting","version":1},"reference_index":89,"source":"pdf_text","source_observed_at":"2026-07-12T08:12:18.373103Z"},"links":{"citing_paper":"/paper/2607.02637"},"observation_digest":"sha256:fbbdd6763deed0802ff607536d21323340c999be6e1ffc9f0897fa6a99500a71","observation_id":"b54a60b9-2173-4784-b63e-38191796bdd3","resolution":{"observed_at":"2026-07-12T08:12:18.373103Z","resolver_source":null,"status":"unresolved"},"standing_notice":{"events":[],"reason":"canonical_work_link_unavailable","source_receipts":[],"state":"unavailable"}}],"paper":{"arxiv_id":"2607.02637","last_updated":"2026-07-02T15:34:46Z","latest_version":1,"primary_category":"cs.LG","snapshot_observed_at":"2026-08-02T18:06:42.028681Z","submitted_at":"2026-07-02T15:34:46Z","title":"Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting"},"reference_resolution":{"displayed":89,"state_counts":{"malformed_identifier":1,"metadata_mismatch":0,"parse_uncertain":0,"unresolved":87,"verified_exact":1,"verified_fuzzy":0},"total_outbound_references":89},"refusal":"A citation records a reference. It does not transfer a finding from one paper to another.","schema":"pith.paper-citation-record.v1","standing_sources":[{"observed_at":"2026-08-07T06:34:17.273281+00:00","source":"crossref"},{"observed_at":"2026-08-07T06:34:11.927384+00:00","source":"retraction_watch"}],"thesis":"As of 7 August 2026, this Paper Citation Record lists 89 of 89 outbound references and 0 inbound Pith citation observations for arXiv:2607.02637."}