{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:OFNTJFKIG6XUTEWYYMIXBIGZNC","short_pith_number":"pith:OFNTJFKI","canonical_record":{"source":{"id":"2605.17938","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:52:26Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"c6aaeaf9d2fcf50ed924f3a0d4115c85e1e254528aa6f07c3b2cbdfade7e66ce","abstract_canon_sha256":"bc564170a2bb0be74ee779ed6d4fabcce2a8b885d5c05d72f279f4c96c8593cb"},"schema_version":"1.0"},"canonical_sha256":"715b34954837af4992d8c31170a0d968a27b5ef8929e06addd54b4697cd2419d","source":{"kind":"arxiv","id":"2605.17938","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17938","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17938v1","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17938","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_12","alias_value":"OFNTJFKIG6XU","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_16","alias_value":"OFNTJFKIG6XUTEWY","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_8","alias_value":"OFNTJFKI","created_at":"2026-05-20T00:05:07Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:OFNTJFKIG6XUTEWYYMIXBIGZNC","target":"record","payload":{"canonical_record":{"source":{"id":"2605.17938","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:52:26Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"c6aaeaf9d2fcf50ed924f3a0d4115c85e1e254528aa6f07c3b2cbdfade7e66ce","abstract_canon_sha256":"bc564170a2bb0be74ee779ed6d4fabcce2a8b885d5c05d72f279f4c96c8593cb"},"schema_version":"1.0"},"canonical_sha256":"715b34954837af4992d8c31170a0d968a27b5ef8929e06addd54b4697cd2419d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:05:07.038056Z","signature_b64":"vI5nZmT8zuibBvg/vu+ndYEbux9JwegNbPso+nrZTQ7JM++XMCphn+lvc4i7MkeKELyzRhLLLaZkD7DelW2aCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"715b34954837af4992d8c31170a0d968a27b5ef8929e06addd54b4697cd2419d","last_reissued_at":"2026-05-20T00:05:07.037178Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:05:07.037178Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.17938","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":"2026-05-20T00:05:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0onUDF4rhv4tyebZJ+++sj0FUC+kctywLsGHEfPEOPabTVWU3W3DT5I59dA6ruJVoj7jZzuq0xMJHk9BSzdGDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T15:36:24.777601Z"},"content_sha256":"fb918a2cc061e0c96a9719b339fef85aed08b3cab440f2b58b1eacbb912d2320","schema_version":"1.0","event_id":"sha256:fb918a2cc061e0c96a9719b339fef85aed08b3cab440f2b58b1eacbb912d2320"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:OFNTJFKIG6XUTEWYYMIXBIGZNC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Training data attribution in diffusion models via mirrored unlearning and noise-consistent skew","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Dipam Goswami, Fabio Morreale, Joan Serr\\`a, Wei-Hsiang Liao, Yuki Mitsufuji","submitted_at":"2026-05-18T06:52:26Z","abstract_excerpt":"Training data attribution (TDA) should enable generative model interpretability and foster a variety of related downstream tasks. Nonetheless, current TDA approaches lack reliability and robustness, preventing their adoption in real-world setups. In this paper, we take a decisive step towards more reliable and robust TDA for diffusion models. We propose to perform TDA with mirrored unlearning and noise-consistent skew (MUCS). The idea is to fine-tune a second model with bounded mirrored gradient ascent, and to measure the normalized skew of this model with respect to the original one using con"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17938","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/2605.17938/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":"2026-05-20T00:05:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xTLtFz4O1CHjwMrRoN9d7i5fCEgwGKTjdhiil5dBU9SFQWGEmJUJP89ZwvNyrY2/GbI9CJTleWpfrT/l7XwyCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-23T15:36:24.778403Z"},"content_sha256":"9a304ec26dd827545a0e64d438bd2ec6f93a00e7abf995575a3a23761b44e780","schema_version":"1.0","event_id":"sha256:9a304ec26dd827545a0e64d438bd2ec6f93a00e7abf995575a3a23761b44e780"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/bundle.json","state_url":"https://pith.science/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/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":"2026-05-23T15:36:24Z","links":{"resolver":"https://pith.science/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC","bundle":"https://pith.science/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/bundle.json","state":"https://pith.science/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OFNTJFKIG6XUTEWYYMIXBIGZNC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:OFNTJFKIG6XUTEWYYMIXBIGZNC","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":"bc564170a2bb0be74ee779ed6d4fabcce2a8b885d5c05d72f279f4c96c8593cb","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:52:26Z","title_canon_sha256":"c6aaeaf9d2fcf50ed924f3a0d4115c85e1e254528aa6f07c3b2cbdfade7e66ce"},"schema_version":"1.0","source":{"id":"2605.17938","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.17938","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"arxiv_version","alias_value":"2605.17938v1","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17938","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_12","alias_value":"OFNTJFKIG6XU","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_16","alias_value":"OFNTJFKIG6XUTEWY","created_at":"2026-05-20T00:05:07Z"},{"alias_kind":"pith_short_8","alias_value":"OFNTJFKI","created_at":"2026-05-20T00:05:07Z"}],"graph_snapshots":[{"event_id":"sha256:9a304ec26dd827545a0e64d438bd2ec6f93a00e7abf995575a3a23761b44e780","target":"graph","created_at":"2026-05-20T00:05:07Z","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/2605.17938/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Training data attribution (TDA) should enable generative model interpretability and foster a variety of related downstream tasks. Nonetheless, current TDA approaches lack reliability and robustness, preventing their adoption in real-world setups. In this paper, we take a decisive step towards more reliable and robust TDA for diffusion models. We propose to perform TDA with mirrored unlearning and noise-consistent skew (MUCS). The idea is to fine-tune a second model with bounded mirrored gradient ascent, and to measure the normalized skew of this model with respect to the original one using con","authors_text":"Dipam Goswami, Fabio Morreale, Joan Serr\\`a, Wei-Hsiang Liao, Yuki Mitsufuji","cross_cats":["cs.AI","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:52:26Z","title":"Training data attribution in diffusion models via mirrored unlearning and noise-consistent skew"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17938","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:fb918a2cc061e0c96a9719b339fef85aed08b3cab440f2b58b1eacbb912d2320","target":"record","created_at":"2026-05-20T00:05:07Z","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":"bc564170a2bb0be74ee779ed6d4fabcce2a8b885d5c05d72f279f4c96c8593cb","cross_cats_sorted":["cs.AI","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-18T06:52:26Z","title_canon_sha256":"c6aaeaf9d2fcf50ed924f3a0d4115c85e1e254528aa6f07c3b2cbdfade7e66ce"},"schema_version":"1.0","source":{"id":"2605.17938","kind":"arxiv","version":1}},"canonical_sha256":"715b34954837af4992d8c31170a0d968a27b5ef8929e06addd54b4697cd2419d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"715b34954837af4992d8c31170a0d968a27b5ef8929e06addd54b4697cd2419d","first_computed_at":"2026-05-20T00:05:07.037178Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T00:05:07.037178Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vI5nZmT8zuibBvg/vu+ndYEbux9JwegNbPso+nrZTQ7JM++XMCphn+lvc4i7MkeKELyzRhLLLaZkD7DelW2aCA==","signature_status":"signed_v1","signed_at":"2026-05-20T00:05:07.038056Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.17938","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fb918a2cc061e0c96a9719b339fef85aed08b3cab440f2b58b1eacbb912d2320","sha256:9a304ec26dd827545a0e64d438bd2ec6f93a00e7abf995575a3a23761b44e780"],"state_sha256":"e02da7c50a031358f8356f897c02ebeea413aa57f5d63c66784ec9397448e07f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zBbHgt8gb1VmCDvrlyM3u7LpGRw+ot/1g9+1j9ZZRfaiCGNMNfPOuyLt9/Ml03uz17k5ny+Yx0AtuUomt3teAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-23T15:36:24.782726Z","bundle_sha256":"d85e18e0f4ad2a8cb3aa2a280ca787f0dc28e5cadce9d87bf13b9c10c7c99127"}}