{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:6JPDOEJOYRA5FEWUO6ZI7PUYDL","short_pith_number":"pith:6JPDOEJO","schema_version":"1.0","canonical_sha256":"f25e37112ec441d292d477b28fbe981adf8a7d1ee6b2778c7e2f73f6bfc0a106","source":{"kind":"arxiv","id":"2601.06162","version":4},"attestation_state":"computed","paper":{"title":"Forget Many, Forget Right: Scalable and Precise Concept Unlearning in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"ScaPre gives diffusion models a closed-form way to unlearn many concepts at once by isolating relevant parameters and stabilizing updates.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Bo Hui, Gen Li, Kaiyuan Deng, Xiaolong Ma, Yang Xiao","submitted_at":"2026-01-06T23:59:17Z","abstract_excerpt":"Text-to-image diffusion models have achieved remarkable progress, yet their use raises copyright and misuse concerns, prompting research into machine unlearning. However, extending multi-concept unlearning to large-scale scenarios remains difficult due to three challenges: (i) conflicting weight updates that hinder unlearning or degrade generation; (ii) imprecise mechanisms that cause collateral damage to similar content; and (iii) reliance on additional data or modules, creating scalability bottlenecks. To address these, we propose Scalable-Precise Concept Unlearning (ScaPre), a unified frame"},"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":"2601.06162","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-01-06T23:59:17Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"b85f83a57e5451235104f6799118ead66deada61c9c148502c67ba2381756b11","abstract_canon_sha256":"1a465136b825769debdf97ba9de6b47e9d0cbfec3dd49e71583602ab791d87f9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:04:20.755499Z","signature_b64":"8kY8ZiU8fbpDI2I/KrUOf4qojAkjrhM4G5S8Bhw63VKRhObL7uovvL1bsxWREIg+DjlzX5k8TrjMDOJzJsq/Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f25e37112ec441d292d477b28fbe981adf8a7d1ee6b2778c7e2f73f6bfc0a106","last_reissued_at":"2026-05-20T00:04:20.754650Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:04:20.754650Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Forget Many, Forget Right: Scalable and Precise Concept Unlearning in Diffusion Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"ScaPre gives diffusion models a closed-form way to unlearn many concepts at once by isolating relevant parameters and stabilizing updates.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Bo Hui, Gen Li, Kaiyuan Deng, Xiaolong Ma, Yang Xiao","submitted_at":"2026-01-06T23:59:17Z","abstract_excerpt":"Text-to-image diffusion models have achieved remarkable progress, yet their use raises copyright and misuse concerns, prompting research into machine unlearning. However, extending multi-concept unlearning to large-scale scenarios remains difficult due to three challenges: (i) conflicting weight updates that hinder unlearning or degrade generation; (ii) imprecise mechanisms that cause collateral damage to similar content; and (iii) reliance on additional data or modules, creating scalability bottlenecks. To address these, we propose Scalable-Precise Concept Unlearning (ScaPre), a unified frame"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ScaPre yields an efficient closed-form solution without requiring auxiliary data or sub-models. It forgets up to ×5 more concepts than the best baseline within acceptable quality limits, achieving state-of-the-art precision and efficiency for large-scale unlearning.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the Informax Decoupler can accurately identify concept-relevant parameters and strictly confine updates to the target subspace without collateral damage, and that spectral trace regularization plus geometry alignment fully resolve conflicting weight updates across large numbers of concepts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ScaPre delivers a closed-form method for precise large-scale multi-concept unlearning in diffusion models that removes up to five times more concepts than prior baselines while preserving image quality.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"ScaPre gives diffusion models a closed-form way to unlearn many concepts at once by isolating relevant parameters and stabilizing updates.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f3335f694d1ac3bbf15e63ce3c732f1fdfc9f22a79f4b32816266f864b39445f"},"source":{"id":"2601.06162","kind":"arxiv","version":4},"verdict":{"id":"74e3634b-c99d-4554-80c6-79251625fb09","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T16:56:24.335078Z","strongest_claim":"ScaPre yields an efficient closed-form solution without requiring auxiliary data or sub-models. It forgets up to ×5 more concepts than the best baseline within acceptable quality limits, achieving state-of-the-art precision and efficiency for large-scale unlearning.","one_line_summary":"ScaPre delivers a closed-form method for precise large-scale multi-concept unlearning in diffusion models that removes up to five times more concepts than prior baselines while preserving image quality.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the Informax Decoupler can accurately identify concept-relevant parameters and strictly confine updates to the target subspace without collateral damage, and that spectral trace regularization plus geometry alignment fully resolve conflicting weight updates across large numbers of concepts.","pith_extraction_headline":"ScaPre gives diffusion models a closed-form way to unlearn many concepts at once by isolating relevant parameters and stabilizing updates."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2601.06162/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":"2601.06162","created_at":"2026-05-20T00:04:20.754776+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.06162v4","created_at":"2026-05-20T00:04:20.754776+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.06162","created_at":"2026-05-20T00:04:20.754776+00:00"},{"alias_kind":"pith_short_12","alias_value":"6JPDOEJOYRA5","created_at":"2026-05-20T00:04:20.754776+00:00"},{"alias_kind":"pith_short_16","alias_value":"6JPDOEJOYRA5FEWU","created_at":"2026-05-20T00:04:20.754776+00:00"},{"alias_kind":"pith_short_8","alias_value":"6JPDOEJO","created_at":"2026-05-20T00:04:20.754776+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.15737","citing_title":"BARRIER: Bounded Activation Regions for Robust Information Erasure","ref_index":10,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL","json":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL.json","graph_json":"https://pith.science/api/pith-number/6JPDOEJOYRA5FEWUO6ZI7PUYDL/graph.json","events_json":"https://pith.science/api/pith-number/6JPDOEJOYRA5FEWUO6ZI7PUYDL/events.json","paper":"https://pith.science/paper/6JPDOEJO"},"agent_actions":{"view_html":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL","download_json":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL.json","view_paper":"https://pith.science/paper/6JPDOEJO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.06162&json=true","fetch_graph":"https://pith.science/api/pith-number/6JPDOEJOYRA5FEWUO6ZI7PUYDL/graph.json","fetch_events":"https://pith.science/api/pith-number/6JPDOEJOYRA5FEWUO6ZI7PUYDL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL/action/storage_attestation","attest_author":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL/action/author_attestation","sign_citation":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL/action/citation_signature","submit_replication":"https://pith.science/pith/6JPDOEJOYRA5FEWUO6ZI7PUYDL/action/replication_record"}},"created_at":"2026-05-20T00:04:20.754776+00:00","updated_at":"2026-05-20T00:04:20.754776+00:00"}