{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LMVVVY6FB6RBROCDHDUTBFHB3R","short_pith_number":"pith:LMVVVY6F","schema_version":"1.0","canonical_sha256":"5b2b5ae3c50fa218b84338e93094e1dc68d80374aa201d076b947c452c760a49","source":{"kind":"arxiv","id":"2602.07120","version":2},"attestation_state":"computed","paper":{"title":"Anchored Decoding: Provably Reducing Copyright Risk for Any Language Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jacqueline He, Jonathan Hayase, Luke Zettlemoyer, Pang Wei Koh, Sewoong Oh, Wen-tau Yih","submitted_at":"2026-02-06T19:00:14Z","abstract_excerpt":"Language models (LMs) tend to memorize portions of their training data and emit verbatim spans. When the underlying sources are sensitive or copyright-protected, such reproduction raises issues of consent and compensation for creators and compliance risks for developers. We propose Anchored Decoding, a plug-and-play inference-time method for suppressing verbatim copying: it enables decoding from any risky LM trained on mixed-license data by keeping generation in bounded proximity to a permissively trained safe LM. Anchored Decoding adaptively allocates a user-chosen information budget over the"},"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":"2602.07120","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-02-06T19:00:14Z","cross_cats_sorted":[],"title_canon_sha256":"6dce6e69d3f8276bc23ed7e730133510259ebeabd7d0e950dee4bf31398689f0","abstract_canon_sha256":"339d57503aa2ae7ed0953e9cafabe49d911ee02931f5125743d618475061edab"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:04:56.013693Z","signature_b64":"X3HZaP/gK6BF5cVlsFbYt+PmrpV7h2cf7wF3ZangIGGGlb1xUeg5TjtqLGvGiqhyxqnuN2mOVi9RhIhYdlC6Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b2b5ae3c50fa218b84338e93094e1dc68d80374aa201d076b947c452c760a49","last_reissued_at":"2026-05-27T01:04:56.012833Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:04:56.012833Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Anchored Decoding: Provably Reducing Copyright Risk for Any Language Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jacqueline He, Jonathan Hayase, Luke Zettlemoyer, Pang Wei Koh, Sewoong Oh, Wen-tau Yih","submitted_at":"2026-02-06T19:00:14Z","abstract_excerpt":"Language models (LMs) tend to memorize portions of their training data and emit verbatim spans. When the underlying sources are sensitive or copyright-protected, such reproduction raises issues of consent and compensation for creators and compliance risks for developers. We propose Anchored Decoding, a plug-and-play inference-time method for suppressing verbatim copying: it enables decoding from any risky LM trained on mixed-license data by keeping generation in bounded proximity to a permissively trained safe LM. Anchored Decoding adaptively allocates a user-chosen information budget over the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.07120","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/2602.07120/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":"2602.07120","created_at":"2026-05-27T01:04:56.012932+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.07120v2","created_at":"2026-05-27T01:04:56.012932+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.07120","created_at":"2026-05-27T01:04:56.012932+00:00"},{"alias_kind":"pith_short_12","alias_value":"LMVVVY6FB6RB","created_at":"2026-05-27T01:04:56.012932+00:00"},{"alias_kind":"pith_short_16","alias_value":"LMVVVY6FB6RBROCD","created_at":"2026-05-27T01:04:56.012932+00:00"},{"alias_kind":"pith_short_8","alias_value":"LMVVVY6F","created_at":"2026-05-27T01:04:56.012932+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/LMVVVY6FB6RBROCDHDUTBFHB3R","json":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R.json","graph_json":"https://pith.science/api/pith-number/LMVVVY6FB6RBROCDHDUTBFHB3R/graph.json","events_json":"https://pith.science/api/pith-number/LMVVVY6FB6RBROCDHDUTBFHB3R/events.json","paper":"https://pith.science/paper/LMVVVY6F"},"agent_actions":{"view_html":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R","download_json":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R.json","view_paper":"https://pith.science/paper/LMVVVY6F","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.07120&json=true","fetch_graph":"https://pith.science/api/pith-number/LMVVVY6FB6RBROCDHDUTBFHB3R/graph.json","fetch_events":"https://pith.science/api/pith-number/LMVVVY6FB6RBROCDHDUTBFHB3R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R/action/storage_attestation","attest_author":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R/action/author_attestation","sign_citation":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R/action/citation_signature","submit_replication":"https://pith.science/pith/LMVVVY6FB6RBROCDHDUTBFHB3R/action/replication_record"}},"created_at":"2026-05-27T01:04:56.012932+00:00","updated_at":"2026-05-27T01:04:56.012932+00:00"}