{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:3AMVYPD52YTZW4HEZA66FNYWB3","short_pith_number":"pith:3AMVYPD5","schema_version":"1.0","canonical_sha256":"d8195c3c7dd6279b70e4c83de2b7160ec5459174a4e34512e859204ff3a90272","source":{"kind":"arxiv","id":"1806.03198","version":3},"attestation_state":"computed","paper":{"title":"Spreading vectors for similarity search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Alexandre Sablayrolles, Cordelia Schmid, Herv\\'e J\\'egou, Matthijs Douze","submitted_at":"2018-06-08T14:46:22Z","abstract_excerpt":"Discretizing multi-dimensional data distributions is a fundamental step of modern indexing methods. State-of-the-art techniques learn parameters of quantizers on training data for optimal performance, thus adapting quantizers to the data. In this work, we propose to reverse this paradigm and adapt the data to the quantizer: we train a neural net which last layer forms a fixed parameter-free quantizer, such as pre-defined points of a hyper-sphere. As a proxy objective, we design and train a neural network that favors uniformity in the spherical latent space, while preserving the neighborhood st"},"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":"1806.03198","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2018-06-08T14:46:22Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d0f229413607b608806c58cf564073c6b0663395f4d82b1a44d5143ac249ee4f","abstract_canon_sha256":"21c5f19bd9fe5632a9c2fd11c53ee3c9be0190fc73f7fec52e104146da7c201b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:00:43.148436Z","signature_b64":"tEk1zVvibkKjwa645g23Qapu9IpiIipUG4/hTHbbIaLEr0hFN1p+PdXXZlLv123HJJmT4OKEvz4Bc8c1QNc2BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d8195c3c7dd6279b70e4c83de2b7160ec5459174a4e34512e859204ff3a90272","last_reissued_at":"2026-07-05T00:00:43.148000Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:00:43.148000Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Spreading vectors for similarity search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Alexandre Sablayrolles, Cordelia Schmid, Herv\\'e J\\'egou, Matthijs Douze","submitted_at":"2018-06-08T14:46:22Z","abstract_excerpt":"Discretizing multi-dimensional data distributions is a fundamental step of modern indexing methods. State-of-the-art techniques learn parameters of quantizers on training data for optimal performance, thus adapting quantizers to the data. In this work, we propose to reverse this paradigm and adapt the data to the quantizer: we train a neural net which last layer forms a fixed parameter-free quantizer, such as pre-defined points of a hyper-sphere. As a proxy objective, we design and train a neural network that favors uniformity in the spherical latent space, while preserving the neighborhood st"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1806.03198","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/1806.03198/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":"1806.03198","created_at":"2026-07-05T00:00:43.148061+00:00"},{"alias_kind":"arxiv_version","alias_value":"1806.03198v3","created_at":"2026-07-05T00:00:43.148061+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1806.03198","created_at":"2026-07-05T00:00:43.148061+00:00"},{"alias_kind":"pith_short_12","alias_value":"3AMVYPD52YTZ","created_at":"2026-07-05T00:00:43.148061+00:00"},{"alias_kind":"pith_short_16","alias_value":"3AMVYPD52YTZW4HE","created_at":"2026-07-05T00:00:43.148061+00:00"},{"alias_kind":"pith_short_8","alias_value":"3AMVYPD5","created_at":"2026-07-05T00:00:43.148061+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17603","citing_title":"Expanding SPHERE-JEPA: A Family of Statistical Regularizers for the Hypersphere","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2510.07191","citing_title":"Resolution scaling governs DINOv3 transfer performance in chest radiograph classification","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2511.11232","citing_title":"DoReMi: Bridging 3D Domains via Topology-Aware Domain-Representation Mixture of Experts","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23166","citing_title":"A satellite foundation model for improved wealth monitoring","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20392","citing_title":"Self-supervised pretraining for an iterative image size agnostic vision transformer","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3","json":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3.json","graph_json":"https://pith.science/api/pith-number/3AMVYPD52YTZW4HEZA66FNYWB3/graph.json","events_json":"https://pith.science/api/pith-number/3AMVYPD52YTZW4HEZA66FNYWB3/events.json","paper":"https://pith.science/paper/3AMVYPD5"},"agent_actions":{"view_html":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3","download_json":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3.json","view_paper":"https://pith.science/paper/3AMVYPD5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1806.03198&json=true","fetch_graph":"https://pith.science/api/pith-number/3AMVYPD52YTZW4HEZA66FNYWB3/graph.json","fetch_events":"https://pith.science/api/pith-number/3AMVYPD52YTZW4HEZA66FNYWB3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3/action/storage_attestation","attest_author":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3/action/author_attestation","sign_citation":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3/action/citation_signature","submit_replication":"https://pith.science/pith/3AMVYPD52YTZW4HEZA66FNYWB3/action/replication_record"}},"created_at":"2026-07-05T00:00:43.148061+00:00","updated_at":"2026-07-05T00:00:43.148061+00:00"}