{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:URFNQ5YOSGXFO5ERIDQQ442TWX","short_pith_number":"pith:URFNQ5YO","schema_version":"1.0","canonical_sha256":"a44ad8770e91ae57749140e10e7353b5f17379bebf3db30a9505b673a56d196d","source":{"kind":"arxiv","id":"1711.11157","version":2},"attestation_state":"computed","paper":{"title":"A Semantic Loss Function for Deep Learning with Symbolic Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.LO","stat.ML"],"primary_cat":"cs.AI","authors_text":"Guy Van den Broeck, Jingyi Xu, Tal Friedman, Yitao Liang, Zilu Zhang","submitted_at":"2017-11-29T23:49:55Z","abstract_excerpt":"This paper develops a novel methodology for using symbolic knowledge in deep learning. From first principles, we derive a semantic loss function that bridges between neural output vectors and logical constraints. This loss function captures how close the neural network is to satisfying the constraints on its output. An experimental evaluation shows that it effectively guides the learner to achieve (near-)state-of-the-art results on semi-supervised multi-class classification. Moreover, it significantly increases the ability of the neural network to predict structured objects, such as rankings a"},"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":"1711.11157","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2017-11-29T23:49:55Z","cross_cats_sorted":["cs.LG","cs.LO","stat.ML"],"title_canon_sha256":"e1749324a5facac59f813601807eec648fdf504d46e23ac2b39392765611eeb5","abstract_canon_sha256":"9ba192db44a0e1f41bf4691c1e30c51b936c3e68dc61419feebe745bb1ad775b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:13:55.114121Z","signature_b64":"F7mfO0eUnr8PdypIsN3exbbTyoJiBNIDQeYHV844EIFsszA1okGYKmHgE4Z+eGr0+/MtWTizUUueMu2QHg4QDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a44ad8770e91ae57749140e10e7353b5f17379bebf3db30a9505b673a56d196d","last_reissued_at":"2026-05-18T00:13:55.113441Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:13:55.113441Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Semantic Loss Function for Deep Learning with Symbolic Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.LO","stat.ML"],"primary_cat":"cs.AI","authors_text":"Guy Van den Broeck, Jingyi Xu, Tal Friedman, Yitao Liang, Zilu Zhang","submitted_at":"2017-11-29T23:49:55Z","abstract_excerpt":"This paper develops a novel methodology for using symbolic knowledge in deep learning. From first principles, we derive a semantic loss function that bridges between neural output vectors and logical constraints. This loss function captures how close the neural network is to satisfying the constraints on its output. An experimental evaluation shows that it effectively guides the learner to achieve (near-)state-of-the-art results on semi-supervised multi-class classification. Moreover, it significantly increases the ability of the neural network to predict structured objects, such as rankings a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1711.11157","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":"1711.11157","created_at":"2026-05-18T00:13:55.113541+00:00"},{"alias_kind":"arxiv_version","alias_value":"1711.11157v2","created_at":"2026-05-18T00:13:55.113541+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1711.11157","created_at":"2026-05-18T00:13:55.113541+00:00"},{"alias_kind":"pith_short_12","alias_value":"URFNQ5YOSGXF","created_at":"2026-05-18T12:31:49.984773+00:00"},{"alias_kind":"pith_short_16","alias_value":"URFNQ5YOSGXFO5ER","created_at":"2026-05-18T12:31:49.984773+00:00"},{"alias_kind":"pith_short_8","alias_value":"URFNQ5YO","created_at":"2026-05-18T12:31:49.984773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2602.09317","citing_title":"SnareNet: Flexible Repair Layers for Neural Networks with Hard Constraints","ref_index":40,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13845","citing_title":"Quantitative Linear Logic for Neuro-Symbolic Learning and Verification","ref_index":131,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13845","citing_title":"Quantitative Linear Logic for Neuro-Symbolic Learning and Verification","ref_index":131,"is_internal_anchor":true},{"citing_arxiv_id":"2604.14360","citing_title":"Digital Guardians: The Past and The Future of Cyber-Physical Resilience","ref_index":241,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX","json":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX.json","graph_json":"https://pith.science/api/pith-number/URFNQ5YOSGXFO5ERIDQQ442TWX/graph.json","events_json":"https://pith.science/api/pith-number/URFNQ5YOSGXFO5ERIDQQ442TWX/events.json","paper":"https://pith.science/paper/URFNQ5YO"},"agent_actions":{"view_html":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX","download_json":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX.json","view_paper":"https://pith.science/paper/URFNQ5YO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1711.11157&json=true","fetch_graph":"https://pith.science/api/pith-number/URFNQ5YOSGXFO5ERIDQQ442TWX/graph.json","fetch_events":"https://pith.science/api/pith-number/URFNQ5YOSGXFO5ERIDQQ442TWX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX/action/storage_attestation","attest_author":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX/action/author_attestation","sign_citation":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX/action/citation_signature","submit_replication":"https://pith.science/pith/URFNQ5YOSGXFO5ERIDQQ442TWX/action/replication_record"}},"created_at":"2026-05-18T00:13:55.113541+00:00","updated_at":"2026-05-18T00:13:55.113541+00:00"}