{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:7RJGSCNFPTLMHAKOFABMH5JA2U","short_pith_number":"pith:7RJGSCNF","schema_version":"1.0","canonical_sha256":"fc526909a57cd6c3814e2802c3f520d50fb283c39ca0dbe4dff6cfe92b39a228","source":{"kind":"arxiv","id":"1802.07595","version":1},"attestation_state":"computed","paper":{"title":"Smooth Loss Functions for Deep Top-k Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrew Zisserman, Leonard Berrada, M. Pawan Kumar","submitted_at":"2018-02-21T14:54:43Z","abstract_excerpt":"The top-k error is a common measure of performance in machine learning and computer vision. In practice, top-k classification is typically performed with deep neural networks trained with the cross-entropy loss. Theoretical results indeed suggest that cross-entropy is an optimal learning objective for such a task in the limit of infinite data. In the context of limited and noisy data however, the use of a loss function that is specifically designed for top-k classification can bring significant improvements. Our empirical evidence suggests that the loss function must be smooth and have non-spa"},"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":"1802.07595","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-02-21T14:54:43Z","cross_cats_sorted":[],"title_canon_sha256":"3aee514bd0ae7bd98305ca9f0857c396a6c697eedc073139f48c317cfcddf391","abstract_canon_sha256":"52d30d36cb953769b326972d6ee385e105a1dfa34417ca90b0b2beb10de2f530"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:22:49.838815Z","signature_b64":"RG+B+Xe6N8fISwXZxfi1pjWVBc1hAX4KRgoEKrOhG98vToO/C2T7T8indmvNjRZUlXry0ZlZg4DB9/7tMdQeBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc526909a57cd6c3814e2802c3f520d50fb283c39ca0dbe4dff6cfe92b39a228","last_reissued_at":"2026-05-18T00:22:49.838343Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:22:49.838343Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Smooth Loss Functions for Deep Top-k Classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrew Zisserman, Leonard Berrada, M. Pawan Kumar","submitted_at":"2018-02-21T14:54:43Z","abstract_excerpt":"The top-k error is a common measure of performance in machine learning and computer vision. In practice, top-k classification is typically performed with deep neural networks trained with the cross-entropy loss. Theoretical results indeed suggest that cross-entropy is an optimal learning objective for such a task in the limit of infinite data. In the context of limited and noisy data however, the use of a loss function that is specifically designed for top-k classification can bring significant improvements. Our empirical evidence suggests that the loss function must be smooth and have non-spa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1802.07595","kind":"arxiv","version":1},"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":"1802.07595","created_at":"2026-05-18T00:22:49.838415+00:00"},{"alias_kind":"arxiv_version","alias_value":"1802.07595v1","created_at":"2026-05-18T00:22:49.838415+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1802.07595","created_at":"2026-05-18T00:22:49.838415+00:00"},{"alias_kind":"pith_short_12","alias_value":"7RJGSCNFPTLM","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_16","alias_value":"7RJGSCNFPTLMHAKO","created_at":"2026-05-18T12:32:11.075285+00:00"},{"alias_kind":"pith_short_8","alias_value":"7RJGSCNF","created_at":"2026-05-18T12:32:11.075285+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.06265","citing_title":"ConquerNet: Convolution-Smoothed Quantile ReLU Neural Networks with Minimax Guarantees","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U","json":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U.json","graph_json":"https://pith.science/api/pith-number/7RJGSCNFPTLMHAKOFABMH5JA2U/graph.json","events_json":"https://pith.science/api/pith-number/7RJGSCNFPTLMHAKOFABMH5JA2U/events.json","paper":"https://pith.science/paper/7RJGSCNF"},"agent_actions":{"view_html":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U","download_json":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U.json","view_paper":"https://pith.science/paper/7RJGSCNF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1802.07595&json=true","fetch_graph":"https://pith.science/api/pith-number/7RJGSCNFPTLMHAKOFABMH5JA2U/graph.json","fetch_events":"https://pith.science/api/pith-number/7RJGSCNFPTLMHAKOFABMH5JA2U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U/action/storage_attestation","attest_author":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U/action/author_attestation","sign_citation":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U/action/citation_signature","submit_replication":"https://pith.science/pith/7RJGSCNFPTLMHAKOFABMH5JA2U/action/replication_record"}},"created_at":"2026-05-18T00:22:49.838415+00:00","updated_at":"2026-05-18T00:22:49.838415+00:00"}