{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:YBVJYVFTLERX43DWEXQZRGIJB5","short_pith_number":"pith:YBVJYVFT","schema_version":"1.0","canonical_sha256":"c06a9c54b359237e6c7625e19899090f5648dcdb800acd919a9a1fec272945f5","source":{"kind":"arxiv","id":"2311.07682","version":2},"attestation_state":"computed","paper":{"title":"Fuse to Forget: Bias Reduction and Selective Memorization through Model Fusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kerem Zaman, Leshem Choshen, Shashank Srivastava","submitted_at":"2023-11-13T19:02:56Z","abstract_excerpt":"Model fusion research aims to aggregate the knowledge of multiple individual models to enhance performance by combining their weights. In this work, we study the inverse problem: investigating whether model fusion can be used to reduce unwanted knowledge. We investigate the effects of model fusion in three scenarios: the learning of shortcuts, social biases, and memorization of training data in fine-tuned language models. Through experiments covering classification and generation tasks, our analysis highlights that shared knowledge among models is enhanced during model fusion, while unshared k"},"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":"2311.07682","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-11-13T19:02:56Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"ffc28b9dacae9839b3aaa1a43244b498da79564786b680a73981f1b5f45f62c0","abstract_canon_sha256":"8baedb82ced193cf036fd7c44a0caab8be946e54fa410acf53bd0e19a181c16a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:18:21.325191Z","signature_b64":"+CNW4ETjosHC0Pg86Uw/JWBZJcTwpuJxJpESiqPSwQHZhtLk5J/XYX5r0huzY2ebTzJUADbDSVs+qvDlCPuQCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c06a9c54b359237e6c7625e19899090f5648dcdb800acd919a9a1fec272945f5","last_reissued_at":"2026-07-05T09:18:21.324747Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:18:21.324747Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fuse to Forget: Bias Reduction and Selective Memorization through Model Fusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kerem Zaman, Leshem Choshen, Shashank Srivastava","submitted_at":"2023-11-13T19:02:56Z","abstract_excerpt":"Model fusion research aims to aggregate the knowledge of multiple individual models to enhance performance by combining their weights. In this work, we study the inverse problem: investigating whether model fusion can be used to reduce unwanted knowledge. We investigate the effects of model fusion in three scenarios: the learning of shortcuts, social biases, and memorization of training data in fine-tuned language models. Through experiments covering classification and generation tasks, our analysis highlights that shared knowledge among models is enhanced during model fusion, while unshared k"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.07682","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/2311.07682/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":"2311.07682","created_at":"2026-07-05T09:18:21.324799+00:00"},{"alias_kind":"arxiv_version","alias_value":"2311.07682v2","created_at":"2026-07-05T09:18:21.324799+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.07682","created_at":"2026-07-05T09:18:21.324799+00:00"},{"alias_kind":"pith_short_12","alias_value":"YBVJYVFTLERX","created_at":"2026-07-05T09:18:21.324799+00:00"},{"alias_kind":"pith_short_16","alias_value":"YBVJYVFTLERX43DW","created_at":"2026-07-05T09:18:21.324799+00:00"},{"alias_kind":"pith_short_8","alias_value":"YBVJYVFT","created_at":"2026-07-05T09:18:21.324799+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2408.07666","citing_title":"Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities","ref_index":284,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5","json":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5.json","graph_json":"https://pith.science/api/pith-number/YBVJYVFTLERX43DWEXQZRGIJB5/graph.json","events_json":"https://pith.science/api/pith-number/YBVJYVFTLERX43DWEXQZRGIJB5/events.json","paper":"https://pith.science/paper/YBVJYVFT"},"agent_actions":{"view_html":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5","download_json":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5.json","view_paper":"https://pith.science/paper/YBVJYVFT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2311.07682&json=true","fetch_graph":"https://pith.science/api/pith-number/YBVJYVFTLERX43DWEXQZRGIJB5/graph.json","fetch_events":"https://pith.science/api/pith-number/YBVJYVFTLERX43DWEXQZRGIJB5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5/action/storage_attestation","attest_author":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5/action/author_attestation","sign_citation":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5/action/citation_signature","submit_replication":"https://pith.science/pith/YBVJYVFTLERX43DWEXQZRGIJB5/action/replication_record"}},"created_at":"2026-07-05T09:18:21.324799+00:00","updated_at":"2026-07-05T09:18:21.324799+00:00"}