{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:KKYGGBHQBCKL7LXO2BZLHNZKF5","short_pith_number":"pith:KKYGGBHQ","schema_version":"1.0","canonical_sha256":"52b06304f00894bfaeeed072b3b72a2f6aea2501261d69962e163d17b98b43fd","source":{"kind":"arxiv","id":"2506.18747","version":1},"attestation_state":"computed","paper":{"title":"ContinualFlow: Learning and Unlearning with Neural Flow Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Davide Bacciu, Lorenzo Simone, Shuangge Ma","submitted_at":"2025-06-23T15:20:58Z","abstract_excerpt":"We introduce ContinualFlow, a principled framework for targeted unlearning in generative models via Flow Matching. Our method leverages an energy-based reweighting loss to softly subtract undesired regions of the data distribution without retraining from scratch or requiring direct access to the samples to be unlearned. Instead, it relies on energy-based proxies to guide the unlearning process. We prove that this induces gradients equivalent to Flow Matching toward a soft mass-subtracted target, and validate the framework through experiments on 2D and image domains, supported by interpretable "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2506.18747","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-23T15:20:58Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1aba44e3ce5e6b7a7c3bcce207325e3fd7c1ff609b4d29b4c107396d0bb0fddb","abstract_canon_sha256":"b11a302eae40996483a2077e97144744db1aa888f3db0593d9c86f74a55cb2b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:55.364482Z","signature_b64":"Wy7CE/8aYm8mHsDkT3M5WMkoJmFugFBP/WCMNmLPm9b22e7v2c1oeoQxmGwICDK1i4UAJlntmi+vdv/tigfBDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"52b06304f00894bfaeeed072b3b72a2f6aea2501261d69962e163d17b98b43fd","last_reissued_at":"2026-07-05T11:25:55.363997Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:55.363997Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ContinualFlow: Learning and Unlearning with Neural Flow Matching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Davide Bacciu, Lorenzo Simone, Shuangge Ma","submitted_at":"2025-06-23T15:20:58Z","abstract_excerpt":"We introduce ContinualFlow, a principled framework for targeted unlearning in generative models via Flow Matching. Our method leverages an energy-based reweighting loss to softly subtract undesired regions of the data distribution without retraining from scratch or requiring direct access to the samples to be unlearned. Instead, it relies on energy-based proxies to guide the unlearning process. We prove that this induces gradients equivalent to Flow Matching toward a soft mass-subtracted target, and validate the framework through experiments on 2D and image domains, supported by interpretable "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.18747","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/2506.18747/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2506.18747","created_at":"2026-07-05T11:25:55.364052+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.18747v1","created_at":"2026-07-05T11:25:55.364052+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.18747","created_at":"2026-07-05T11:25:55.364052+00:00"},{"alias_kind":"pith_short_12","alias_value":"KKYGGBHQBCKL","created_at":"2026-07-05T11:25:55.364052+00:00"},{"alias_kind":"pith_short_16","alias_value":"KKYGGBHQBCKL7LXO","created_at":"2026-07-05T11:25:55.364052+00:00"},{"alias_kind":"pith_short_8","alias_value":"KKYGGBHQ","created_at":"2026-07-05T11:25:55.364052+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5","json":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5.json","graph_json":"https://pith.science/api/pith-number/KKYGGBHQBCKL7LXO2BZLHNZKF5/graph.json","events_json":"https://pith.science/api/pith-number/KKYGGBHQBCKL7LXO2BZLHNZKF5/events.json","paper":"https://pith.science/paper/KKYGGBHQ"},"agent_actions":{"view_html":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5","download_json":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5.json","view_paper":"https://pith.science/paper/KKYGGBHQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.18747&json=true","fetch_graph":"https://pith.science/api/pith-number/KKYGGBHQBCKL7LXO2BZLHNZKF5/graph.json","fetch_events":"https://pith.science/api/pith-number/KKYGGBHQBCKL7LXO2BZLHNZKF5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5/action/storage_attestation","attest_author":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5/action/author_attestation","sign_citation":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5/action/citation_signature","submit_replication":"https://pith.science/pith/KKYGGBHQBCKL7LXO2BZLHNZKF5/action/replication_record"}},"created_at":"2026-07-05T11:25:55.364052+00:00","updated_at":"2026-07-05T11:25:55.364052+00:00"}