{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:3UF7DYTAMZZKUMBECHBSQ4EVE7","short_pith_number":"pith:3UF7DYTA","schema_version":"1.0","canonical_sha256":"dd0bf1e2606672aa302411c328709527c633a78faf7abff29c8f7a33496cd09c","source":{"kind":"arxiv","id":"1911.02855","version":3},"attestation_state":"computed","paper":{"title":"Dice Loss for Data-imbalanced NLP Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fei Wu, Jiwei Li, Junjun Liang, Xiaofei Sun, Xiaoya Li, Yuxian Meng","submitted_at":"2019-11-07T11:14:05Z","abstract_excerpt":"Many NLP tasks such as tagging and machine reading comprehension are faced with the severe data imbalance issue: negative examples significantly outnumber positive examples, and the huge number of background examples (or easy-negative examples) overwhelms the training. The most commonly used cross entropy (CE) criteria is actually an accuracy-oriented objective, and thus creates a discrepancy between training and test: at training time, each training instance contributes equally to the objective function, while at test time F1 score concerns more about positive examples. In this paper, we prop"},"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":"1911.02855","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-11-07T11:14:05Z","cross_cats_sorted":[],"title_canon_sha256":"8cf120384ea1cbb8debe7da8370dff482daa923f5620b60fb33cdc64dc20a125","abstract_canon_sha256":"b45a6d8cbb5896a2fafbae2c81a79a3e84e2ddd07acfc4ce82b9147248834a44"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:31:22.740217Z","signature_b64":"PfW+zD77MOUzpOzIwPUuXtjLyf6y+/y+kG5ya6ZfX4aSPII9XerSG7T+R1UdjWwSn+tTJIIN8V9V2XkNGouQCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dd0bf1e2606672aa302411c328709527c633a78faf7abff29c8f7a33496cd09c","last_reissued_at":"2026-07-05T01:31:22.739728Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:31:22.739728Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dice Loss for Data-imbalanced NLP Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fei Wu, Jiwei Li, Junjun Liang, Xiaofei Sun, Xiaoya Li, Yuxian Meng","submitted_at":"2019-11-07T11:14:05Z","abstract_excerpt":"Many NLP tasks such as tagging and machine reading comprehension are faced with the severe data imbalance issue: negative examples significantly outnumber positive examples, and the huge number of background examples (or easy-negative examples) overwhelms the training. The most commonly used cross entropy (CE) criteria is actually an accuracy-oriented objective, and thus creates a discrepancy between training and test: at training time, each training instance contributes equally to the objective function, while at test time F1 score concerns more about positive examples. In this paper, we prop"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1911.02855","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1911.02855/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"1911.02855","created_at":"2026-07-05T01:31:22.739785+00:00"},{"alias_kind":"arxiv_version","alias_value":"1911.02855v3","created_at":"2026-07-05T01:31:22.739785+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1911.02855","created_at":"2026-07-05T01:31:22.739785+00:00"},{"alias_kind":"pith_short_12","alias_value":"3UF7DYTAMZZK","created_at":"2026-07-05T01:31:22.739785+00:00"},{"alias_kind":"pith_short_16","alias_value":"3UF7DYTAMZZKUMBE","created_at":"2026-07-05T01:31:22.739785+00:00"},{"alias_kind":"pith_short_8","alias_value":"3UF7DYTA","created_at":"2026-07-05T01:31:22.739785+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2409.11022","citing_title":"DynamicNER: A Dynamic, Multilingual, and Fine-Grained Dataset for LLM-based Named Entity Recognition","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2604.04632","citing_title":"InCTRLv2: Generalist Residual Models for Few-Shot Anomaly Detection and Segmentation","ref_index":30,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7","json":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7.json","graph_json":"https://pith.science/api/pith-number/3UF7DYTAMZZKUMBECHBSQ4EVE7/graph.json","events_json":"https://pith.science/api/pith-number/3UF7DYTAMZZKUMBECHBSQ4EVE7/events.json","paper":"https://pith.science/paper/3UF7DYTA"},"agent_actions":{"view_html":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7","download_json":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7.json","view_paper":"https://pith.science/paper/3UF7DYTA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1911.02855&json=true","fetch_graph":"https://pith.science/api/pith-number/3UF7DYTAMZZKUMBECHBSQ4EVE7/graph.json","fetch_events":"https://pith.science/api/pith-number/3UF7DYTAMZZKUMBECHBSQ4EVE7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7/action/storage_attestation","attest_author":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7/action/author_attestation","sign_citation":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7/action/citation_signature","submit_replication":"https://pith.science/pith/3UF7DYTAMZZKUMBECHBSQ4EVE7/action/replication_record"}},"created_at":"2026-07-05T01:31:22.739785+00:00","updated_at":"2026-07-05T01:31:22.739785+00:00"}