{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:4YIDV4QP5WASVAHXGDAML6CSXT","short_pith_number":"pith:4YIDV4QP","schema_version":"1.0","canonical_sha256":"e6103af20fed812a80f730c0c5f852bce72eeb522595e3d6ebaaf37aab19b67c","source":{"kind":"arxiv","id":"1605.09774","version":2},"attestation_state":"computed","paper":{"title":"Asynchrony begets Momentum, with an Application to Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Ce Zhang, Christopher R\\'e, Ioannis Mitliagkas, Stefan Hadjis","submitted_at":"2016-05-31T19:16:56Z","abstract_excerpt":"Asynchronous methods are widely used in deep learning, but have limited theoretical justification when applied to non-convex problems. We show that running stochastic gradient descent (SGD) in an asynchronous manner can be viewed as adding a momentum-like term to the SGD iteration. Our result does not assume convexity of the objective function, so it is applicable to deep learning systems. We observe that a standard queuing model of asynchrony results in a form of momentum that is commonly used by deep learning practitioners. This forges a link between queuing theory and asynchrony in deep lea"},"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":"1605.09774","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2016-05-31T19:16:56Z","cross_cats_sorted":["cs.DC","cs.LG","math.OC"],"title_canon_sha256":"176d14c79e8dcc3bd41c70a4827c906234520ac994d7f99b0ee86f4ecdefd4a6","abstract_canon_sha256":"c66c5b33b9089dbab0d06edd5573605a0412da9f957ec29eb321af2a6f40f6fb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:56:46.056765Z","signature_b64":"oby6+VnJZcSghHVnYmjJNOkTBqUjmZL4bDRyspyvffLB1bkN8Z3uHQBD7xQY/KrO2D9PKUROq1YB/py2WipnAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e6103af20fed812a80f730c0c5f852bce72eeb522595e3d6ebaaf37aab19b67c","last_reissued_at":"2026-05-18T00:56:46.056188Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:56:46.056188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Asynchrony begets Momentum, with an Application to Deep Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Ce Zhang, Christopher R\\'e, Ioannis Mitliagkas, Stefan Hadjis","submitted_at":"2016-05-31T19:16:56Z","abstract_excerpt":"Asynchronous methods are widely used in deep learning, but have limited theoretical justification when applied to non-convex problems. We show that running stochastic gradient descent (SGD) in an asynchronous manner can be viewed as adding a momentum-like term to the SGD iteration. Our result does not assume convexity of the objective function, so it is applicable to deep learning systems. We observe that a standard queuing model of asynchrony results in a form of momentum that is commonly used by deep learning practitioners. This forges a link between queuing theory and asynchrony in deep lea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1605.09774","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":""},"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":"1605.09774","created_at":"2026-05-18T00:56:46.056301+00:00"},{"alias_kind":"arxiv_version","alias_value":"1605.09774v2","created_at":"2026-05-18T00:56:46.056301+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1605.09774","created_at":"2026-05-18T00:56:46.056301+00:00"},{"alias_kind":"pith_short_12","alias_value":"4YIDV4QP5WAS","created_at":"2026-05-18T12:29:58.707656+00:00"},{"alias_kind":"pith_short_16","alias_value":"4YIDV4QP5WASVAHX","created_at":"2026-05-18T12:29:58.707656+00:00"},{"alias_kind":"pith_short_8","alias_value":"4YIDV4QP","created_at":"2026-05-18T12:29:58.707656+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.14108","citing_title":"Momentum Further Constrains Sharpness at the Edge of Stochastic Stability","ref_index":33,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT","json":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT.json","graph_json":"https://pith.science/api/pith-number/4YIDV4QP5WASVAHXGDAML6CSXT/graph.json","events_json":"https://pith.science/api/pith-number/4YIDV4QP5WASVAHXGDAML6CSXT/events.json","paper":"https://pith.science/paper/4YIDV4QP"},"agent_actions":{"view_html":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT","download_json":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT.json","view_paper":"https://pith.science/paper/4YIDV4QP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1605.09774&json=true","fetch_graph":"https://pith.science/api/pith-number/4YIDV4QP5WASVAHXGDAML6CSXT/graph.json","fetch_events":"https://pith.science/api/pith-number/4YIDV4QP5WASVAHXGDAML6CSXT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT/action/storage_attestation","attest_author":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT/action/author_attestation","sign_citation":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT/action/citation_signature","submit_replication":"https://pith.science/pith/4YIDV4QP5WASVAHXGDAML6CSXT/action/replication_record"}},"created_at":"2026-05-18T00:56:46.056301+00:00","updated_at":"2026-05-18T00:56:46.056301+00:00"}