{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:B4DOITX7CFMOD264EPWI7D7N6I","short_pith_number":"pith:B4DOITX7","schema_version":"1.0","canonical_sha256":"0f06e44eff1158e1ebdc23ec8f8fedf22bd436ab29102c544de2a264aadb7377","source":{"kind":"arxiv","id":"2111.09839","version":1},"attestation_state":"computed","paper":{"title":"Training Neural Networks with Fixed Sparse Masks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Colin Raffel, Varun Nair, Yi-Lin Sung","submitted_at":"2021-11-18T18:06:01Z","abstract_excerpt":"During typical gradient-based training of deep neural networks, all of the model's parameters are updated at each iteration. Recent work has shown that it is possible to update only a small subset of the model's parameters during training, which can alleviate storage and communication requirements. In this paper, we show that it is possible to induce a fixed sparse mask on the model's parameters that selects a subset to update over many iterations. Our method constructs the mask out of the $k$ parameters with the largest Fisher information as a simple approximation as to which parameters are m"},"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":"2111.09839","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-11-18T18:06:01Z","cross_cats_sorted":[],"title_canon_sha256":"71ab3c551b1c2e8e8043f36d2bd8b5ee76b515cc38aaa58e7a85a308872f45b3","abstract_canon_sha256":"4815076d77f9f3f7623714953c01bb3bd4bc7a015ef548cd76db22a3e0a6bfe8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:33:16.202476Z","signature_b64":"ojANFhmwxUKq1n786PTp97xghDxXUrAOU8J0twtIlGyuHXf3JAlJd7i+EgSW/1l5enl6BAvCm0Xens1zuz/MCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0f06e44eff1158e1ebdc23ec8f8fedf22bd436ab29102c544de2a264aadb7377","last_reissued_at":"2026-07-05T03:33:16.202080Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:33:16.202080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Neural Networks with Fixed Sparse Masks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Colin Raffel, Varun Nair, Yi-Lin Sung","submitted_at":"2021-11-18T18:06:01Z","abstract_excerpt":"During typical gradient-based training of deep neural networks, all of the model's parameters are updated at each iteration. Recent work has shown that it is possible to update only a small subset of the model's parameters during training, which can alleviate storage and communication requirements. In this paper, we show that it is possible to induce a fixed sparse mask on the model's parameters that selects a subset to update over many iterations. Our method constructs the mask out of the $k$ parameters with the largest Fisher information as a simple approximation as to which parameters are m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2111.09839","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/2111.09839/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":"2111.09839","created_at":"2026-07-05T03:33:16.202138+00:00"},{"alias_kind":"arxiv_version","alias_value":"2111.09839v1","created_at":"2026-07-05T03:33:16.202138+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2111.09839","created_at":"2026-07-05T03:33:16.202138+00:00"},{"alias_kind":"pith_short_12","alias_value":"B4DOITX7CFMO","created_at":"2026-07-05T03:33:16.202138+00:00"},{"alias_kind":"pith_short_16","alias_value":"B4DOITX7CFMOD264","created_at":"2026-07-05T03:33:16.202138+00:00"},{"alias_kind":"pith_short_8","alias_value":"B4DOITX7","created_at":"2026-07-05T03:33:16.202138+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2212.04089","citing_title":"Editing Models with Task Arithmetic","ref_index":96,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I","json":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I.json","graph_json":"https://pith.science/api/pith-number/B4DOITX7CFMOD264EPWI7D7N6I/graph.json","events_json":"https://pith.science/api/pith-number/B4DOITX7CFMOD264EPWI7D7N6I/events.json","paper":"https://pith.science/paper/B4DOITX7"},"agent_actions":{"view_html":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I","download_json":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I.json","view_paper":"https://pith.science/paper/B4DOITX7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2111.09839&json=true","fetch_graph":"https://pith.science/api/pith-number/B4DOITX7CFMOD264EPWI7D7N6I/graph.json","fetch_events":"https://pith.science/api/pith-number/B4DOITX7CFMOD264EPWI7D7N6I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I/action/storage_attestation","attest_author":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I/action/author_attestation","sign_citation":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I/action/citation_signature","submit_replication":"https://pith.science/pith/B4DOITX7CFMOD264EPWI7D7N6I/action/replication_record"}},"created_at":"2026-07-05T03:33:16.202138+00:00","updated_at":"2026-07-05T03:33:16.202138+00:00"}