{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:4265TCZXZ7QUVZHNVJC2NMAE3W","short_pith_number":"pith:4265TCZX","schema_version":"1.0","canonical_sha256":"e6bdd98b37cfe14ae4edaa45a6b004ddac4ddb8015d8e53f63e6a0faf8dffd1d","source":{"kind":"arxiv","id":"1704.04451","version":3},"attestation_state":"computed","paper":{"title":"Optimizing Differentiable Relaxations of Coreference Evaluation Metrics","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ivan Titov, Phong Le","submitted_at":"2017-04-14T15:22:51Z","abstract_excerpt":"Coreference evaluation metrics are hard to optimize directly as they are non-differentiable functions, not easily decomposable into elementary decisions. Consequently, most approaches optimize objectives only indirectly related to the end goal, resulting in suboptimal performance. Instead, we propose a differentiable relaxation that lends itself to gradient-based optimisation, thus bypassing the need for reinforcement learning or heuristic modification of cross-entropy. We show that by modifying the training objective of a competitive neural coreference system, we obtain a substantial gain in "},"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":"1704.04451","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2017-04-14T15:22:51Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"0aa9854abf53cdac4df12256b934d9fddc1ff97dc60b9858a0f6924cb82f3820","abstract_canon_sha256":"8ddfcedc9d990ffc90c0763b62181da58f673b845d20de4a9178cd3d68b1e910"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:41:52.987202Z","signature_b64":"7Lt/634sCXNvQ9icx3j85dGwNWbzkXaC4zMkH3HGpsqVAkd5iwvIC7z4oouocddOCHaZNNO7g4nEOzk5PwIMCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e6bdd98b37cfe14ae4edaa45a6b004ddac4ddb8015d8e53f63e6a0faf8dffd1d","last_reissued_at":"2026-05-18T00:41:52.986755Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:41:52.986755Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Optimizing Differentiable Relaxations of Coreference Evaluation Metrics","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ivan Titov, Phong Le","submitted_at":"2017-04-14T15:22:51Z","abstract_excerpt":"Coreference evaluation metrics are hard to optimize directly as they are non-differentiable functions, not easily decomposable into elementary decisions. Consequently, most approaches optimize objectives only indirectly related to the end goal, resulting in suboptimal performance. Instead, we propose a differentiable relaxation that lends itself to gradient-based optimisation, thus bypassing the need for reinforcement learning or heuristic modification of cross-entropy. We show that by modifying the training objective of a competitive neural coreference system, we obtain a substantial gain in "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1704.04451","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":"1704.04451","created_at":"2026-05-18T00:41:52.986823+00:00"},{"alias_kind":"arxiv_version","alias_value":"1704.04451v3","created_at":"2026-05-18T00:41:52.986823+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1704.04451","created_at":"2026-05-18T00:41:52.986823+00:00"},{"alias_kind":"pith_short_12","alias_value":"4265TCZXZ7QU","created_at":"2026-05-18T12:30:58.224056+00:00"},{"alias_kind":"pith_short_16","alias_value":"4265TCZXZ7QUVZHN","created_at":"2026-05-18T12:30:58.224056+00:00"},{"alias_kind":"pith_short_8","alias_value":"4265TCZX","created_at":"2026-05-18T12:30:58.224056+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/4265TCZXZ7QUVZHNVJC2NMAE3W","json":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W.json","graph_json":"https://pith.science/api/pith-number/4265TCZXZ7QUVZHNVJC2NMAE3W/graph.json","events_json":"https://pith.science/api/pith-number/4265TCZXZ7QUVZHNVJC2NMAE3W/events.json","paper":"https://pith.science/paper/4265TCZX"},"agent_actions":{"view_html":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W","download_json":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W.json","view_paper":"https://pith.science/paper/4265TCZX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1704.04451&json=true","fetch_graph":"https://pith.science/api/pith-number/4265TCZXZ7QUVZHNVJC2NMAE3W/graph.json","fetch_events":"https://pith.science/api/pith-number/4265TCZXZ7QUVZHNVJC2NMAE3W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W/action/storage_attestation","attest_author":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W/action/author_attestation","sign_citation":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W/action/citation_signature","submit_replication":"https://pith.science/pith/4265TCZXZ7QUVZHNVJC2NMAE3W/action/replication_record"}},"created_at":"2026-05-18T00:41:52.986823+00:00","updated_at":"2026-05-18T00:41:52.986823+00:00"}