{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:6OXUEVBRN7ZORO53QNDX2YRM5G","short_pith_number":"pith:6OXUEVBR","canonical_record":{"source":{"id":"2505.09768","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-14T19:54:55Z","cross_cats_sorted":[],"title_canon_sha256":"7c87296d47b6e1c1489d47d0bf426363709e883ad30be82347b3d3370ae019e6","abstract_canon_sha256":"2b316a05e18907550e50d7eb9d3c73e99685c8256b29a189466a0e9a44865f7e"},"schema_version":"1.0"},"canonical_sha256":"f3af4254316ff2e8bbbb83477d622ce9ad7c64c425dc84f48feac155e5ed3674","source":{"kind":"arxiv","id":"2505.09768","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.09768","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.09768v1","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.09768","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_12","alias_value":"6OXUEVBRN7ZO","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_16","alias_value":"6OXUEVBRN7ZORO53","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_8","alias_value":"6OXUEVBR","created_at":"2026-07-05T11:03:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:6OXUEVBRN7ZORO53QNDX2YRM5G","target":"record","payload":{"canonical_record":{"source":{"id":"2505.09768","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-14T19:54:55Z","cross_cats_sorted":[],"title_canon_sha256":"7c87296d47b6e1c1489d47d0bf426363709e883ad30be82347b3d3370ae019e6","abstract_canon_sha256":"2b316a05e18907550e50d7eb9d3c73e99685c8256b29a189466a0e9a44865f7e"},"schema_version":"1.0"},"canonical_sha256":"f3af4254316ff2e8bbbb83477d622ce9ad7c64c425dc84f48feac155e5ed3674","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:03:28.202234Z","signature_b64":"SR93waMN2G4DMAHQDxALvvoZX44Es48YZG6QNQpZU9F+XEFVTnyb37hlx0VUEpfA+K5emi9BgzuOWQYID2fIBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f3af4254316ff2e8bbbb83477d622ce9ad7c64c425dc84f48feac155e5ed3674","last_reissued_at":"2026-07-05T11:03:28.201765Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:03:28.201765Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.09768","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:03:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gsIcRtlBdu2w1pXELBshdevougqxv7Yzq3scI2zGH5FIxCUm8mczsrpKFQ+YqZPSPRQW8paD7C/1nG0jHjiuBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T08:45:09.696348Z"},"content_sha256":"c95e7cb1134ad5afddab2232002b07848a26d8eb48a36e276e33a3301b2eb9b5","schema_version":"1.0","event_id":"sha256:c95e7cb1134ad5afddab2232002b07848a26d8eb48a36e276e33a3301b2eb9b5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:6OXUEVBRN7ZORO53QNDX2YRM5G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Self-Consuming Generative Models with Adversarially Curated Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Xiukun Wei, Xueru Zhang","submitted_at":"2025-05-14T19:54:55Z","abstract_excerpt":"Recent advances in generative models have made it increasingly difficult to distinguish real data from model-generated synthetic data. Using synthetic data for successive training of future model generations creates \"self-consuming loops\", which may lead to model collapse or training instability. Furthermore, synthetic data is often subject to human feedback and curated by users based on their preferences. Ferbach et al. (2024) recently showed that when data is curated according to user preferences, the self-consuming retraining loop drives the model to converge toward a distribution that opti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.09768","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2505.09768/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:03:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"w7hnCK4yMlXsDGExk7smwI42p2c6/GVTV+mGZKjumalnSJ1Sb1d4f1AcLnW5CtDvDK8GuXFGW0ftq/+QJ0NDAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T08:45:09.696896Z"},"content_sha256":"74b23fcb96f27480fb6be7cedb2156d1110f9cffed20d01307e5ce11a19f7982","schema_version":"1.0","event_id":"sha256:74b23fcb96f27480fb6be7cedb2156d1110f9cffed20d01307e5ce11a19f7982"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/bundle.json","state_url":"https://pith.science/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-18T08:45:09Z","links":{"resolver":"https://pith.science/pith/6OXUEVBRN7ZORO53QNDX2YRM5G","bundle":"https://pith.science/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/bundle.json","state":"https://pith.science/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6OXUEVBRN7ZORO53QNDX2YRM5G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6OXUEVBRN7ZORO53QNDX2YRM5G","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2b316a05e18907550e50d7eb9d3c73e99685c8256b29a189466a0e9a44865f7e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-14T19:54:55Z","title_canon_sha256":"7c87296d47b6e1c1489d47d0bf426363709e883ad30be82347b3d3370ae019e6"},"schema_version":"1.0","source":{"id":"2505.09768","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.09768","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.09768v1","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.09768","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_12","alias_value":"6OXUEVBRN7ZO","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_16","alias_value":"6OXUEVBRN7ZORO53","created_at":"2026-07-05T11:03:28Z"},{"alias_kind":"pith_short_8","alias_value":"6OXUEVBR","created_at":"2026-07-05T11:03:28Z"}],"graph_snapshots":[{"event_id":"sha256:74b23fcb96f27480fb6be7cedb2156d1110f9cffed20d01307e5ce11a19f7982","target":"graph","created_at":"2026-07-05T11:03:28Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2505.09768/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in generative models have made it increasingly difficult to distinguish real data from model-generated synthetic data. Using synthetic data for successive training of future model generations creates \"self-consuming loops\", which may lead to model collapse or training instability. Furthermore, synthetic data is often subject to human feedback and curated by users based on their preferences. Ferbach et al. (2024) recently showed that when data is curated according to user preferences, the self-consuming retraining loop drives the model to converge toward a distribution that opti","authors_text":"Xiukun Wei, Xueru Zhang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-14T19:54:55Z","title":"Self-Consuming Generative Models with Adversarially Curated Data"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.09768","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c95e7cb1134ad5afddab2232002b07848a26d8eb48a36e276e33a3301b2eb9b5","target":"record","created_at":"2026-07-05T11:03:28Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"2b316a05e18907550e50d7eb9d3c73e99685c8256b29a189466a0e9a44865f7e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-14T19:54:55Z","title_canon_sha256":"7c87296d47b6e1c1489d47d0bf426363709e883ad30be82347b3d3370ae019e6"},"schema_version":"1.0","source":{"id":"2505.09768","kind":"arxiv","version":1}},"canonical_sha256":"f3af4254316ff2e8bbbb83477d622ce9ad7c64c425dc84f48feac155e5ed3674","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f3af4254316ff2e8bbbb83477d622ce9ad7c64c425dc84f48feac155e5ed3674","first_computed_at":"2026-07-05T11:03:28.201765Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:03:28.201765Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SR93waMN2G4DMAHQDxALvvoZX44Es48YZG6QNQpZU9F+XEFVTnyb37hlx0VUEpfA+K5emi9BgzuOWQYID2fIBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:03:28.202234Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.09768","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c95e7cb1134ad5afddab2232002b07848a26d8eb48a36e276e33a3301b2eb9b5","sha256:74b23fcb96f27480fb6be7cedb2156d1110f9cffed20d01307e5ce11a19f7982"],"state_sha256":"1abd08606243ea3ed64a42b1a5ac1eeb5e00bf97b4230f71c7ad7070a81ba45e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"W90hFSV2etaP5jeRwm4HyYANF5PVOLE7Q51Jzqh1jvMd58LyNaA03HSHsfXNqbwajMN6z8q+S7+dIJy34zGNDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T08:45:09.700740Z","bundle_sha256":"deea2c1601f671762b53a99b09e32621b97fa36c34ff1c9579c318266ed41074"}}