{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:TRF3ZVA4VYPZY2UFAVDXYD3N7F","short_pith_number":"pith:TRF3ZVA4","canonical_record":{"source":{"id":"2503.13769","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T23:17:16Z","cross_cats_sorted":[],"title_canon_sha256":"73c451f200bff2d0dd12059b0de36ae840ee7d4127a01abd0bf3f04922ef5c7a","abstract_canon_sha256":"085e03d0b28156ac49fa31ef5f3eb8ab54aba7c89be54d50bb8c539d19070923"},"schema_version":"1.0"},"canonical_sha256":"9c4bbcd41cae1f9c6a8505477c0f6df95d0f9aac97ded45fce3d228656f71b12","source":{"kind":"arxiv","id":"2503.13769","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.13769","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"arxiv_version","alias_value":"2503.13769v2","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13769","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_12","alias_value":"TRF3ZVA4VYPZ","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_16","alias_value":"TRF3ZVA4VYPZY2UF","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_8","alias_value":"TRF3ZVA4","created_at":"2026-07-05T10:37:07Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:TRF3ZVA4VYPZY2UFAVDXYD3N7F","target":"record","payload":{"canonical_record":{"source":{"id":"2503.13769","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T23:17:16Z","cross_cats_sorted":[],"title_canon_sha256":"73c451f200bff2d0dd12059b0de36ae840ee7d4127a01abd0bf3f04922ef5c7a","abstract_canon_sha256":"085e03d0b28156ac49fa31ef5f3eb8ab54aba7c89be54d50bb8c539d19070923"},"schema_version":"1.0"},"canonical_sha256":"9c4bbcd41cae1f9c6a8505477c0f6df95d0f9aac97ded45fce3d228656f71b12","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:37:07.799880Z","signature_b64":"mJ/kBgUm+bBe/SfxkR4EDG/suvfoKi4XtAnj96futOY2kg0ryN4qCLMsTCS5mU3ojsSeLyQOSWNdfH5I+TlrAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9c4bbcd41cae1f9c6a8505477c0f6df95d0f9aac97ded45fce3d228656f71b12","last_reissued_at":"2026-07-05T10:37:07.798981Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:37:07.798981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.13769","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:37:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xUYwUnmQv8cDA+V8HX6vXLKBj0WwFX7X8MZJp7RcPqNKZXv8AG/sLAtFbJJ0jYau/k0O9I1C8hOvXO2eJN5BDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T20:56:46.966476Z"},"content_sha256":"bdb4c61396ee9847e861dd1dee1c219c54ceac7719fceb7fcfa869236eab5a3b","schema_version":"1.0","event_id":"sha256:bdb4c61396ee9847e861dd1dee1c219c54ceac7719fceb7fcfa869236eab5a3b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:TRF3ZVA4VYPZY2UFAVDXYD3N7F","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Continual Unlearning for Foundational Text-to-Image Models without Generalization Erosion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Kartik Thakral, Mayank Vatsa, Richa Singh, Tal Hassner, Tamar Glaser","submitted_at":"2025-03-17T23:17:16Z","abstract_excerpt":"How can we effectively unlearn selected concepts from pre-trained generative foundation models without resorting to extensive retraining? This research introduces `continual unlearning', a novel paradigm that enables the targeted removal of multiple specific concepts from foundational generative models, incrementally. We propose Decremental Unlearning without Generalization Erosion (DUGE) algorithm which selectively unlearns the generation of undesired concepts while preserving the generation of related, non-targeted concepts and alleviating generalization erosion. For this, DUGE targets three"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13769","kind":"arxiv","version":2},"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/2503.13769/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":"2026-07-05T10:37:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ptXrIVH2gSzdiovsCBBnv10wSNQVM4H29yZxliWCVsHy+Y0XnvOE9txbclQiCvOJIo1yhF9RH8KMCQs0JToGBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T20:56:46.966986Z"},"content_sha256":"23133ff12278517920c37f4bab8efd97cf7326b5e6fde6dac66bbf50a91eb20d","schema_version":"1.0","event_id":"sha256:23133ff12278517920c37f4bab8efd97cf7326b5e6fde6dac66bbf50a91eb20d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/bundle.json","state_url":"https://pith.science/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/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":"2026-08-08T20:56:46Z","links":{"resolver":"https://pith.science/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F","bundle":"https://pith.science/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/bundle.json","state":"https://pith.science/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TRF3ZVA4VYPZY2UFAVDXYD3N7F/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TRF3ZVA4VYPZY2UFAVDXYD3N7F","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":"085e03d0b28156ac49fa31ef5f3eb8ab54aba7c89be54d50bb8c539d19070923","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T23:17:16Z","title_canon_sha256":"73c451f200bff2d0dd12059b0de36ae840ee7d4127a01abd0bf3f04922ef5c7a"},"schema_version":"1.0","source":{"id":"2503.13769","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.13769","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"arxiv_version","alias_value":"2503.13769v2","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.13769","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_12","alias_value":"TRF3ZVA4VYPZ","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_16","alias_value":"TRF3ZVA4VYPZY2UF","created_at":"2026-07-05T10:37:07Z"},{"alias_kind":"pith_short_8","alias_value":"TRF3ZVA4","created_at":"2026-07-05T10:37:07Z"}],"graph_snapshots":[{"event_id":"sha256:23133ff12278517920c37f4bab8efd97cf7326b5e6fde6dac66bbf50a91eb20d","target":"graph","created_at":"2026-07-05T10:37:07Z","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/2503.13769/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"How can we effectively unlearn selected concepts from pre-trained generative foundation models without resorting to extensive retraining? This research introduces `continual unlearning', a novel paradigm that enables the targeted removal of multiple specific concepts from foundational generative models, incrementally. We propose Decremental Unlearning without Generalization Erosion (DUGE) algorithm which selectively unlearns the generation of undesired concepts while preserving the generation of related, non-targeted concepts and alleviating generalization erosion. For this, DUGE targets three","authors_text":"Kartik Thakral, Mayank Vatsa, Richa Singh, Tal Hassner, Tamar Glaser","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T23:17:16Z","title":"Continual Unlearning for Foundational Text-to-Image Models without Generalization Erosion"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.13769","kind":"arxiv","version":2},"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:bdb4c61396ee9847e861dd1dee1c219c54ceac7719fceb7fcfa869236eab5a3b","target":"record","created_at":"2026-07-05T10:37:07Z","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":"085e03d0b28156ac49fa31ef5f3eb8ab54aba7c89be54d50bb8c539d19070923","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-17T23:17:16Z","title_canon_sha256":"73c451f200bff2d0dd12059b0de36ae840ee7d4127a01abd0bf3f04922ef5c7a"},"schema_version":"1.0","source":{"id":"2503.13769","kind":"arxiv","version":2}},"canonical_sha256":"9c4bbcd41cae1f9c6a8505477c0f6df95d0f9aac97ded45fce3d228656f71b12","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9c4bbcd41cae1f9c6a8505477c0f6df95d0f9aac97ded45fce3d228656f71b12","first_computed_at":"2026-07-05T10:37:07.798981Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:37:07.798981Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mJ/kBgUm+bBe/SfxkR4EDG/suvfoKi4XtAnj96futOY2kg0ryN4qCLMsTCS5mU3ojsSeLyQOSWNdfH5I+TlrAg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:37:07.799880Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.13769","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bdb4c61396ee9847e861dd1dee1c219c54ceac7719fceb7fcfa869236eab5a3b","sha256:23133ff12278517920c37f4bab8efd97cf7326b5e6fde6dac66bbf50a91eb20d"],"state_sha256":"546c1e64a6e79db8958e12741d76d646ae99c6c289978b0aec91b51a925cffb6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rc7CE61nQLYE9tGY1thZuMELtrT1RCcCxhwCBl3N04k3idUZXKsTJH1rbNEP2heg1imlBo5j4s5A04YiZcsKDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T20:56:46.972198Z","bundle_sha256":"710fb7df593edebb6c438a3bdf6ed3762571991a5c1dda516b8cf7080571b284"}}