{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2015:LOBOWILE3MMEYDOKLI2N42JW2Q","short_pith_number":"pith:LOBOWILE","schema_version":"1.0","canonical_sha256":"5b82eb2164db184c0dca5a34de6936d43fe2bb5a973be6656eda1c3ff93026bf","source":{"kind":"arxiv","id":"1511.06411","version":2},"attestation_state":"computed","paper":{"title":"Training Deep Neural Networks via Direct Loss Minimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander G. Schwing, Raquel Urtasun, Richard S. Zemel, Yang Song","submitted_at":"2015-11-19T22:02:26Z","abstract_excerpt":"Supervised training of deep neural nets typically relies on minimizing cross-entropy. However, in many domains, we are interested in performing well on metrics specific to the application. In this paper we propose a direct loss minimization approach to train deep neural networks, which provably minimizes the application-specific loss function. This is often non-trivial, since these functions are neither smooth nor decomposable and thus are not amenable to optimization with standard gradient-based methods. We demonstrate the effectiveness of our approach in the context of maximizing average pre"},"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":"1511.06411","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2015-11-19T22:02:26Z","cross_cats_sorted":[],"title_canon_sha256":"3d2a9565800e0f545e35022201ce48507c99a95dd369cb70897b9ed188f0ea96","abstract_canon_sha256":"ecc73f77fc67d7dc49b209fb938ab2e19bd7ec68132e21affa8c611202a838b3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:13:04.832828Z","signature_b64":"lJRz9oQ7IRmPGLRFxivCmTWGB3Qy/rdHFyud7VtmkscpjS6R+j2Wo0etU0Ld3UkqCQZIfqCTpT+Tu7l2SE+3Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b82eb2164db184c0dca5a34de6936d43fe2bb5a973be6656eda1c3ff93026bf","last_reissued_at":"2026-05-18T01:13:04.832495Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:13:04.832495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Deep Neural Networks via Direct Loss Minimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexander G. Schwing, Raquel Urtasun, Richard S. Zemel, Yang Song","submitted_at":"2015-11-19T22:02:26Z","abstract_excerpt":"Supervised training of deep neural nets typically relies on minimizing cross-entropy. However, in many domains, we are interested in performing well on metrics specific to the application. In this paper we propose a direct loss minimization approach to train deep neural networks, which provably minimizes the application-specific loss function. This is often non-trivial, since these functions are neither smooth nor decomposable and thus are not amenable to optimization with standard gradient-based methods. We demonstrate the effectiveness of our approach in the context of maximizing average pre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1511.06411","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":"1511.06411","created_at":"2026-05-18T01:13:04.832550+00:00"},{"alias_kind":"arxiv_version","alias_value":"1511.06411v2","created_at":"2026-05-18T01:13:04.832550+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1511.06411","created_at":"2026-05-18T01:13:04.832550+00:00"},{"alias_kind":"pith_short_12","alias_value":"LOBOWILE3MME","created_at":"2026-05-18T12:29:29.992203+00:00"},{"alias_kind":"pith_short_16","alias_value":"LOBOWILE3MMEYDOK","created_at":"2026-05-18T12:29:29.992203+00:00"},{"alias_kind":"pith_short_8","alias_value":"LOBOWILE","created_at":"2026-05-18T12:29:29.992203+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/LOBOWILE3MMEYDOKLI2N42JW2Q","json":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q.json","graph_json":"https://pith.science/api/pith-number/LOBOWILE3MMEYDOKLI2N42JW2Q/graph.json","events_json":"https://pith.science/api/pith-number/LOBOWILE3MMEYDOKLI2N42JW2Q/events.json","paper":"https://pith.science/paper/LOBOWILE"},"agent_actions":{"view_html":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q","download_json":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q.json","view_paper":"https://pith.science/paper/LOBOWILE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1511.06411&json=true","fetch_graph":"https://pith.science/api/pith-number/LOBOWILE3MMEYDOKLI2N42JW2Q/graph.json","fetch_events":"https://pith.science/api/pith-number/LOBOWILE3MMEYDOKLI2N42JW2Q/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q/action/storage_attestation","attest_author":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q/action/author_attestation","sign_citation":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q/action/citation_signature","submit_replication":"https://pith.science/pith/LOBOWILE3MMEYDOKLI2N42JW2Q/action/replication_record"}},"created_at":"2026-05-18T01:13:04.832550+00:00","updated_at":"2026-05-18T01:13:04.832550+00:00"}