{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:DY5RCOPB7RYRCCVQDLRVDAEW63","short_pith_number":"pith:DY5RCOPB","schema_version":"1.0","canonical_sha256":"1e3b1139e1fc71110ab01ae3518096f6d7466cf05e72aa8ac1dc4e9a6c5f3fcc","source":{"kind":"arxiv","id":"1706.05699","version":3},"attestation_state":"computed","paper":{"title":"Gradient Diversity: a Key Ingredient for Scalable Distributed Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Ashwin Pananjady, Dimitris Papailiopoulos, Dong Yin, Kannan Ramchandran, Max Lam, Peter Bartlett","submitted_at":"2017-06-18T18:37:12Z","abstract_excerpt":"It has been experimentally observed that distributed implementations of mini-batch stochastic gradient descent (SGD) algorithms exhibit speedup saturation and decaying generalization ability beyond a particular batch-size. In this work, we present an analysis hinting that high similarity between concurrently processed gradients may be a cause of this performance degradation. We introduce the notion of gradient diversity that measures the dissimilarity between concurrent gradient updates, and show its key role in the performance of mini-batch SGD. We prove that on problems with high gradient di"},"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":"1706.05699","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-06-18T18:37:12Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"5293d1717b9fe9a21cbcc018161989135381a9d1dd80ba306e67a2c2fa3224e2","abstract_canon_sha256":"5117252af6895aea63b21cfac88943c17073f9bce6af9696179667dcda746769"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:26:36.921403Z","signature_b64":"wrB3vl8OtREKvDVxsO/oTUMMHILETONMkVOmcs2W3Prtu4kcOQFY30xavMaWZkBAZzUYhLsLx/WFz7fSKVlHCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1e3b1139e1fc71110ab01ae3518096f6d7466cf05e72aa8ac1dc4e9a6c5f3fcc","last_reissued_at":"2026-05-18T00:26:36.919892Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:26:36.919892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Diversity: a Key Ingredient for Scalable Distributed Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Ashwin Pananjady, Dimitris Papailiopoulos, Dong Yin, Kannan Ramchandran, Max Lam, Peter Bartlett","submitted_at":"2017-06-18T18:37:12Z","abstract_excerpt":"It has been experimentally observed that distributed implementations of mini-batch stochastic gradient descent (SGD) algorithms exhibit speedup saturation and decaying generalization ability beyond a particular batch-size. In this work, we present an analysis hinting that high similarity between concurrently processed gradients may be a cause of this performance degradation. We introduce the notion of gradient diversity that measures the dissimilarity between concurrent gradient updates, and show its key role in the performance of mini-batch SGD. We prove that on problems with high gradient di"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1706.05699","kind":"arxiv","version":3},"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":"1706.05699","created_at":"2026-05-18T00:26:36.919959+00:00"},{"alias_kind":"arxiv_version","alias_value":"1706.05699v3","created_at":"2026-05-18T00:26:36.919959+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1706.05699","created_at":"2026-05-18T00:26:36.919959+00:00"},{"alias_kind":"pith_short_12","alias_value":"DY5RCOPB7RYR","created_at":"2026-05-18T12:31:12.930513+00:00"},{"alias_kind":"pith_short_16","alias_value":"DY5RCOPB7RYRCCVQ","created_at":"2026-05-18T12:31:12.930513+00:00"},{"alias_kind":"pith_short_8","alias_value":"DY5RCOPB","created_at":"2026-05-18T12:31:12.930513+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2102.01293","citing_title":"Scaling Laws for Transfer","ref_index":156,"is_internal_anchor":true},{"citing_arxiv_id":"2112.00861","citing_title":"A General Language Assistant as a Laboratory for Alignment","ref_index":198,"is_internal_anchor":false},{"citing_arxiv_id":"2207.05221","citing_title":"Language Models (Mostly) Know What They Know","ref_index":276,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63","json":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63.json","graph_json":"https://pith.science/api/pith-number/DY5RCOPB7RYRCCVQDLRVDAEW63/graph.json","events_json":"https://pith.science/api/pith-number/DY5RCOPB7RYRCCVQDLRVDAEW63/events.json","paper":"https://pith.science/paper/DY5RCOPB"},"agent_actions":{"view_html":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63","download_json":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63.json","view_paper":"https://pith.science/paper/DY5RCOPB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1706.05699&json=true","fetch_graph":"https://pith.science/api/pith-number/DY5RCOPB7RYRCCVQDLRVDAEW63/graph.json","fetch_events":"https://pith.science/api/pith-number/DY5RCOPB7RYRCCVQDLRVDAEW63/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63/action/storage_attestation","attest_author":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63/action/author_attestation","sign_citation":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63/action/citation_signature","submit_replication":"https://pith.science/pith/DY5RCOPB7RYRCCVQDLRVDAEW63/action/replication_record"}},"created_at":"2026-05-18T00:26:36.919959+00:00","updated_at":"2026-05-18T00:26:36.919959+00:00"}