{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:3MYRDF2HK3JORNKEHV6NE55O6R","short_pith_number":"pith:3MYRDF2H","schema_version":"1.0","canonical_sha256":"db3111974756d2e8b5443d7cd277aef469ed963bc9a0fe6d53bc00f2304c40b5","source":{"kind":"arxiv","id":"2304.07288","version":2},"attestation_state":"computed","paper":{"title":"Cross-Entropy Loss Functions: Theoretical Analysis and Applications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anqi Mao, Mehryar Mohri, Yutao Zhong","submitted_at":"2023-04-14T17:58:23Z","abstract_excerpt":"Cross-entropy is a widely used loss function in applications. It coincides with the logistic loss applied to the outputs of a neural network, when the softmax is used. But, what guarantees can we rely on when using cross-entropy as a surrogate loss? We present a theoretical analysis of a broad family of loss functions, comp-sum losses, that includes cross-entropy (or logistic loss), generalized cross-entropy, the mean absolute error and other cross-entropy-like loss functions. We give the first $H$-consistency bounds for these loss functions. These are non-asymptotic guarantees that upper boun"},"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":"2304.07288","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-04-14T17:58:23Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"18e5aed285bfe17a1eaa98901b339e73338b4a8c86f449a4f685d7dca364e1b4","abstract_canon_sha256":"e7e48080decf547665eb527e5a8929cf177f33cf2735d8e24e9ade1bc208eafc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:22:03.118083Z","signature_b64":"X5e8Uc1G3s16S9nD+TYBV3BOIyiL+KGtZpCG+nGfwHHMLQYNAfmU5phMrRZ3az2IABdku4GFnxG/y5SKvsgsBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"db3111974756d2e8b5443d7cd277aef469ed963bc9a0fe6d53bc00f2304c40b5","last_reissued_at":"2026-07-05T06:22:03.117632Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:22:03.117632Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Cross-Entropy Loss Functions: Theoretical Analysis and Applications","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anqi Mao, Mehryar Mohri, Yutao Zhong","submitted_at":"2023-04-14T17:58:23Z","abstract_excerpt":"Cross-entropy is a widely used loss function in applications. It coincides with the logistic loss applied to the outputs of a neural network, when the softmax is used. But, what guarantees can we rely on when using cross-entropy as a surrogate loss? We present a theoretical analysis of a broad family of loss functions, comp-sum losses, that includes cross-entropy (or logistic loss), generalized cross-entropy, the mean absolute error and other cross-entropy-like loss functions. We give the first $H$-consistency bounds for these loss functions. These are non-asymptotic guarantees that upper boun"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.07288","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2304.07288/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":"2304.07288","created_at":"2026-07-05T06:22:03.117688+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.07288v2","created_at":"2026-07-05T06:22:03.117688+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.07288","created_at":"2026-07-05T06:22:03.117688+00:00"},{"alias_kind":"pith_short_12","alias_value":"3MYRDF2HK3JO","created_at":"2026-07-05T06:22:03.117688+00:00"},{"alias_kind":"pith_short_16","alias_value":"3MYRDF2HK3JORNKE","created_at":"2026-07-05T06:22:03.117688+00:00"},{"alias_kind":"pith_short_8","alias_value":"3MYRDF2H","created_at":"2026-07-05T06:22:03.117688+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2410.15761","citing_title":"Optimal Query Allocation in Extractive QA with LLMs: A Learning-to-Defer Framework with Theoretical Guarantees","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2403.07691","citing_title":"ORPO: Monolithic Preference Optimization without Reference Model","ref_index":36,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R","json":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R.json","graph_json":"https://pith.science/api/pith-number/3MYRDF2HK3JORNKEHV6NE55O6R/graph.json","events_json":"https://pith.science/api/pith-number/3MYRDF2HK3JORNKEHV6NE55O6R/events.json","paper":"https://pith.science/paper/3MYRDF2H"},"agent_actions":{"view_html":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R","download_json":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R.json","view_paper":"https://pith.science/paper/3MYRDF2H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.07288&json=true","fetch_graph":"https://pith.science/api/pith-number/3MYRDF2HK3JORNKEHV6NE55O6R/graph.json","fetch_events":"https://pith.science/api/pith-number/3MYRDF2HK3JORNKEHV6NE55O6R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R/action/storage_attestation","attest_author":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R/action/author_attestation","sign_citation":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R/action/citation_signature","submit_replication":"https://pith.science/pith/3MYRDF2HK3JORNKEHV6NE55O6R/action/replication_record"}},"created_at":"2026-07-05T06:22:03.117688+00:00","updated_at":"2026-07-05T06:22:03.117688+00:00"}