{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:RBIKTXNIL5ZNSXJC75TV4KFAWW","short_pith_number":"pith:RBIKTXNI","schema_version":"1.0","canonical_sha256":"8850a9dda85f72d95d22ff675e28a0b5a875afad8d2b1d142eab2bd6682d1e44","source":{"kind":"arxiv","id":"1607.05691","version":1},"attestation_state":"computed","paper":{"title":"Information-theoretical label embeddings for large-scale image classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Fran\\c{c}ois Chollet","submitted_at":"2016-07-19T18:40:01Z","abstract_excerpt":"We present a method for training multi-label, massively multi-class image classification models, that is faster and more accurate than supervision via a sigmoid cross-entropy loss (logistic regression). Our method consists in embedding high-dimensional sparse labels onto a lower-dimensional dense sphere of unit-normed vectors, and treating the classification problem as a cosine proximity regression problem on this sphere. We test our method on a dataset of 300 million high-resolution images with 17,000 labels, where it yields considerably faster convergence, as well as a 7% higher mean average"},"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":"1607.05691","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2016-07-19T18:40:01Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"a34864553c61252b4fffe65a749d51e0641f2385298d734a0f9e9c27611e1530","abstract_canon_sha256":"210c4c41800cbbcb184b092f352d40e44b8f10145cd1202744c9e8c0ab712722"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:10:46.244290Z","signature_b64":"iQjFBo4pC2eE9cr59G3ZzbLo0L2OsaQWnOGAVKzX+kteLORDnz9QmN/ObvqJ+TL9D5QPTy373TMU1H9gMC31Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8850a9dda85f72d95d22ff675e28a0b5a875afad8d2b1d142eab2bd6682d1e44","last_reissued_at":"2026-05-18T01:10:46.243738Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:10:46.243738Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Information-theoretical label embeddings for large-scale image classification","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Fran\\c{c}ois Chollet","submitted_at":"2016-07-19T18:40:01Z","abstract_excerpt":"We present a method for training multi-label, massively multi-class image classification models, that is faster and more accurate than supervision via a sigmoid cross-entropy loss (logistic regression). Our method consists in embedding high-dimensional sparse labels onto a lower-dimensional dense sphere of unit-normed vectors, and treating the classification problem as a cosine proximity regression problem on this sphere. We test our method on a dataset of 300 million high-resolution images with 17,000 labels, where it yields considerably faster convergence, as well as a 7% higher mean average"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1607.05691","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":"1607.05691","created_at":"2026-05-18T01:10:46.243851+00:00"},{"alias_kind":"arxiv_version","alias_value":"1607.05691v1","created_at":"2026-05-18T01:10:46.243851+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1607.05691","created_at":"2026-05-18T01:10:46.243851+00:00"},{"alias_kind":"pith_short_12","alias_value":"RBIKTXNIL5ZN","created_at":"2026-05-18T12:30:41.710351+00:00"},{"alias_kind":"pith_short_16","alias_value":"RBIKTXNIL5ZNSXJC","created_at":"2026-05-18T12:30:41.710351+00:00"},{"alias_kind":"pith_short_8","alias_value":"RBIKTXNI","created_at":"2026-05-18T12:30:41.710351+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/RBIKTXNIL5ZNSXJC75TV4KFAWW","json":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW.json","graph_json":"https://pith.science/api/pith-number/RBIKTXNIL5ZNSXJC75TV4KFAWW/graph.json","events_json":"https://pith.science/api/pith-number/RBIKTXNIL5ZNSXJC75TV4KFAWW/events.json","paper":"https://pith.science/paper/RBIKTXNI"},"agent_actions":{"view_html":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW","download_json":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW.json","view_paper":"https://pith.science/paper/RBIKTXNI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1607.05691&json=true","fetch_graph":"https://pith.science/api/pith-number/RBIKTXNIL5ZNSXJC75TV4KFAWW/graph.json","fetch_events":"https://pith.science/api/pith-number/RBIKTXNIL5ZNSXJC75TV4KFAWW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW/action/storage_attestation","attest_author":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW/action/author_attestation","sign_citation":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW/action/citation_signature","submit_replication":"https://pith.science/pith/RBIKTXNIL5ZNSXJC75TV4KFAWW/action/replication_record"}},"created_at":"2026-05-18T01:10:46.243851+00:00","updated_at":"2026-05-18T01:10:46.243851+00:00"}