{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:2C6N4PJAN7QFVG46RHGU66YAXA","short_pith_number":"pith:2C6N4PJA","schema_version":"1.0","canonical_sha256":"d0bcde3d206fe05a9b9e89cd4f7b00b8344615748f7260a26e2995f4b47602bb","source":{"kind":"arxiv","id":"2010.11305","version":2},"attestation_state":"computed","paper":{"title":"Mixed-Precision Embedding Using a Cache","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Andrew Tulloch, Jianyu Huang, Jie Amy Yang, Jongsoo Park, Ping Tak Peter Tang","submitted_at":"2020-10-21T20:49:54Z","abstract_excerpt":"In recommendation systems, practitioners observed that increase in the number of embedding tables and their sizes often leads to significant improvement in model performances. Given this and the business importance of these models to major internet companies, embedding tables for personalization tasks have grown to terabyte scale and continue to grow at a significant rate. Meanwhile, these large-scale models are often trained with GPUs where high-performance memory is a scarce resource, thus motivating numerous work on embedding table compression during training. We propose a novel change to e"},"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":"2010.11305","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-10-21T20:49:54Z","cross_cats_sorted":["cs.AI","cs.DC"],"title_canon_sha256":"841523daa8a6e99607168875db9cc6e973423667df0f9d596a63cb3eb430c560","abstract_canon_sha256":"8c1ad11a46d1b346804b968ae99bfa893915af1b619dc9eb0512da22e5684bfe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:45:31.835687Z","signature_b64":"y4l4NWa3/COWZCZzzwxUpWxuI10EmKHTjnfNvWEVSX9ngQ6RdPyCxVng7dXz9tQn1JowboaAnvGwTVWFnIb7DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d0bcde3d206fe05a9b9e89cd4f7b00b8344615748f7260a26e2995f4b47602bb","last_reissued_at":"2026-07-05T01:45:31.835234Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:45:31.835234Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mixed-Precision Embedding Using a Cache","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.DC"],"primary_cat":"cs.LG","authors_text":"Andrew Tulloch, Jianyu Huang, Jie Amy Yang, Jongsoo Park, Ping Tak Peter Tang","submitted_at":"2020-10-21T20:49:54Z","abstract_excerpt":"In recommendation systems, practitioners observed that increase in the number of embedding tables and their sizes often leads to significant improvement in model performances. Given this and the business importance of these models to major internet companies, embedding tables for personalization tasks have grown to terabyte scale and continue to grow at a significant rate. Meanwhile, these large-scale models are often trained with GPUs where high-performance memory is a scarce resource, thus motivating numerous work on embedding table compression during training. We propose a novel change to e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.11305","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/2010.11305/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":"2010.11305","created_at":"2026-07-05T01:45:31.835299+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.11305v2","created_at":"2026-07-05T01:45:31.835299+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.11305","created_at":"2026-07-05T01:45:31.835299+00:00"},{"alias_kind":"pith_short_12","alias_value":"2C6N4PJAN7QF","created_at":"2026-07-05T01:45:31.835299+00:00"},{"alias_kind":"pith_short_16","alias_value":"2C6N4PJAN7QFVG46","created_at":"2026-07-05T01:45:31.835299+00:00"},{"alias_kind":"pith_short_8","alias_value":"2C6N4PJA","created_at":"2026-07-05T01:45:31.835299+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2503.09315","citing_title":"ShuffleGate: Scalable Feature Optimization for Recommender Systems via Batch-wise Sensitivity Learning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24073","citing_title":"FreeScale: Distributed Training for Sequence Recommendation Models with Minimal Scaling Cost","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA","json":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA.json","graph_json":"https://pith.science/api/pith-number/2C6N4PJAN7QFVG46RHGU66YAXA/graph.json","events_json":"https://pith.science/api/pith-number/2C6N4PJAN7QFVG46RHGU66YAXA/events.json","paper":"https://pith.science/paper/2C6N4PJA"},"agent_actions":{"view_html":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA","download_json":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA.json","view_paper":"https://pith.science/paper/2C6N4PJA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.11305&json=true","fetch_graph":"https://pith.science/api/pith-number/2C6N4PJAN7QFVG46RHGU66YAXA/graph.json","fetch_events":"https://pith.science/api/pith-number/2C6N4PJAN7QFVG46RHGU66YAXA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA/action/storage_attestation","attest_author":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA/action/author_attestation","sign_citation":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA/action/citation_signature","submit_replication":"https://pith.science/pith/2C6N4PJAN7QFVG46RHGU66YAXA/action/replication_record"}},"created_at":"2026-07-05T01:45:31.835299+00:00","updated_at":"2026-07-05T01:45:31.835299+00:00"}