{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:REE4RIT5ZT4SPJT6SYTQBY3MH7","short_pith_number":"pith:REE4RIT5","schema_version":"1.0","canonical_sha256":"8909c8a27dccf927a67e962700e36c3feaaa3a01b3de5e33a50a30746707ca11","source":{"kind":"arxiv","id":"2106.12379","version":2},"attestation_state":"computed","paper":{"title":"AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adrian Vladu, Alexandra Peste, Dan Alistarh, Eugenia Iofinova","submitted_at":"2021-06-23T13:23:00Z","abstract_excerpt":"The increasing computational requirements of deep neural networks (DNNs) have led to significant interest in obtaining DNN models that are sparse, yet accurate. Recent work has investigated the even harder case of sparse training, where the DNN weights are, for as much as possible, already sparse to reduce computational costs during training. Existing sparse training methods are often empirical and can have lower accuracy relative to the dense baseline. In this paper, we present a general approach called Alternating Compressed/DeCompressed (AC/DC) training of DNNs, demonstrate convergence for "},"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":"2106.12379","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-23T13:23:00Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fbf68c8b69d51ef9ea8f5deda2eaa03714d6c0483dcc7fb48a5745416f557eed","abstract_canon_sha256":"025c184ce27d03a6b8857f8b56a08ee242c5af87bfa16b0370abc3e0d52d3ccd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:41:16.419662Z","signature_b64":"uYvLA/WXdhdlcw7ju9yDTS0bftLfpKadHeFR6F07Qh1QcOKjlZkjRO2cJJgUVwnBpxGIeI5ni71pGnRqJTLeCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8909c8a27dccf927a67e962700e36c3feaaa3a01b3de5e33a50a30746707ca11","last_reissued_at":"2026-07-05T03:41:16.419210Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:41:16.419210Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AC/DC: Alternating Compressed/DeCompressed Training of Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Adrian Vladu, Alexandra Peste, Dan Alistarh, Eugenia Iofinova","submitted_at":"2021-06-23T13:23:00Z","abstract_excerpt":"The increasing computational requirements of deep neural networks (DNNs) have led to significant interest in obtaining DNN models that are sparse, yet accurate. Recent work has investigated the even harder case of sparse training, where the DNN weights are, for as much as possible, already sparse to reduce computational costs during training. Existing sparse training methods are often empirical and can have lower accuracy relative to the dense baseline. In this paper, we present a general approach called Alternating Compressed/DeCompressed (AC/DC) training of DNNs, demonstrate convergence for "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.12379","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/2106.12379/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":"2106.12379","created_at":"2026-07-05T03:41:16.419268+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.12379v2","created_at":"2026-07-05T03:41:16.419268+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.12379","created_at":"2026-07-05T03:41:16.419268+00:00"},{"alias_kind":"pith_short_12","alias_value":"REE4RIT5ZT4S","created_at":"2026-07-05T03:41:16.419268+00:00"},{"alias_kind":"pith_short_16","alias_value":"REE4RIT5ZT4SPJT6","created_at":"2026-07-05T03:41:16.419268+00:00"},{"alias_kind":"pith_short_8","alias_value":"REE4RIT5","created_at":"2026-07-05T03:41:16.419268+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/REE4RIT5ZT4SPJT6SYTQBY3MH7","json":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7.json","graph_json":"https://pith.science/api/pith-number/REE4RIT5ZT4SPJT6SYTQBY3MH7/graph.json","events_json":"https://pith.science/api/pith-number/REE4RIT5ZT4SPJT6SYTQBY3MH7/events.json","paper":"https://pith.science/paper/REE4RIT5"},"agent_actions":{"view_html":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7","download_json":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7.json","view_paper":"https://pith.science/paper/REE4RIT5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.12379&json=true","fetch_graph":"https://pith.science/api/pith-number/REE4RIT5ZT4SPJT6SYTQBY3MH7/graph.json","fetch_events":"https://pith.science/api/pith-number/REE4RIT5ZT4SPJT6SYTQBY3MH7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7/action/storage_attestation","attest_author":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7/action/author_attestation","sign_citation":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7/action/citation_signature","submit_replication":"https://pith.science/pith/REE4RIT5ZT4SPJT6SYTQBY3MH7/action/replication_record"}},"created_at":"2026-07-05T03:41:16.419268+00:00","updated_at":"2026-07-05T03:41:16.419268+00:00"}