{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:H7JBBKAAFBKVYSYIR2LP67T3OF","short_pith_number":"pith:H7JBBKAA","schema_version":"1.0","canonical_sha256":"3fd210a80028555c4b088e96ff7e7b7179cbf2134b140faa6f44e21819cfde5f","source":{"kind":"arxiv","id":"1704.04932","version":2},"attestation_state":"computed","paper":{"title":"Deep Relaxation: partial differential equations for optimizing deep neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.AP","math.OC"],"primary_cat":"cs.LG","authors_text":"Adam Oberman, Guillaume Carlier, Pratik Chaudhari, Stanley Osher, Stefano Soatto","submitted_at":"2017-04-17T11:21:32Z","abstract_excerpt":"In this paper we establish a connection between non-convex optimization methods for training deep neural networks and nonlinear partial differential equations (PDEs). Relaxation techniques arising in statistical physics which have already been used successfully in this context are reinterpreted as solutions of a viscous Hamilton-Jacobi PDE. Using a stochastic control interpretation allows we prove that the modified algorithm performs better in expectation that stochastic gradient descent. Well-known PDE regularity results allow us to analyze the geometry of the relaxed energy landscape, confir"},"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":"1704.04932","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-04-17T11:21:32Z","cross_cats_sorted":["math.AP","math.OC"],"title_canon_sha256":"687b43b95880af43f7df60fda9bc6839555b1990f740c7a937a91501eb66d852","abstract_canon_sha256":"796859d569e4bd2f4ef8bab3071aaed40cfba1c6def02fdd7ee469e1abc6656a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:43:12.989144Z","signature_b64":"CauQo4l+GxVnX7j6wyhpKKFAw3OUoPjRrHwSLSnZLkONTGJz7SY0F03+L2Uz3voJvasun/roInz07OQ4FP0XAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3fd210a80028555c4b088e96ff7e7b7179cbf2134b140faa6f44e21819cfde5f","last_reissued_at":"2026-05-18T00:43:12.988575Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:43:12.988575Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deep Relaxation: partial differential equations for optimizing deep neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["math.AP","math.OC"],"primary_cat":"cs.LG","authors_text":"Adam Oberman, Guillaume Carlier, Pratik Chaudhari, Stanley Osher, Stefano Soatto","submitted_at":"2017-04-17T11:21:32Z","abstract_excerpt":"In this paper we establish a connection between non-convex optimization methods for training deep neural networks and nonlinear partial differential equations (PDEs). Relaxation techniques arising in statistical physics which have already been used successfully in this context are reinterpreted as solutions of a viscous Hamilton-Jacobi PDE. Using a stochastic control interpretation allows we prove that the modified algorithm performs better in expectation that stochastic gradient descent. Well-known PDE regularity results allow us to analyze the geometry of the relaxed energy landscape, confir"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1704.04932","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":"1704.04932","created_at":"2026-05-18T00:43:12.988681+00:00"},{"alias_kind":"arxiv_version","alias_value":"1704.04932v2","created_at":"2026-05-18T00:43:12.988681+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1704.04932","created_at":"2026-05-18T00:43:12.988681+00:00"},{"alias_kind":"pith_short_12","alias_value":"H7JBBKAAFBKV","created_at":"2026-05-18T12:31:18.294218+00:00"},{"alias_kind":"pith_short_16","alias_value":"H7JBBKAAFBKVYSYI","created_at":"2026-05-18T12:31:18.294218+00:00"},{"alias_kind":"pith_short_8","alias_value":"H7JBBKAA","created_at":"2026-05-18T12:31:18.294218+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/H7JBBKAAFBKVYSYIR2LP67T3OF","json":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF.json","graph_json":"https://pith.science/api/pith-number/H7JBBKAAFBKVYSYIR2LP67T3OF/graph.json","events_json":"https://pith.science/api/pith-number/H7JBBKAAFBKVYSYIR2LP67T3OF/events.json","paper":"https://pith.science/paper/H7JBBKAA"},"agent_actions":{"view_html":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF","download_json":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF.json","view_paper":"https://pith.science/paper/H7JBBKAA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1704.04932&json=true","fetch_graph":"https://pith.science/api/pith-number/H7JBBKAAFBKVYSYIR2LP67T3OF/graph.json","fetch_events":"https://pith.science/api/pith-number/H7JBBKAAFBKVYSYIR2LP67T3OF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF/action/storage_attestation","attest_author":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF/action/author_attestation","sign_citation":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF/action/citation_signature","submit_replication":"https://pith.science/pith/H7JBBKAAFBKVYSYIR2LP67T3OF/action/replication_record"}},"created_at":"2026-05-18T00:43:12.988681+00:00","updated_at":"2026-05-18T00:43:12.988681+00:00"}