{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:545AVTSROBWSMKESQQQYL4G5VC","short_pith_number":"pith:545AVTSR","schema_version":"1.0","canonical_sha256":"ef3a0ace51706d262892842185f0dda8869435a05d74dcaa883b5e40364fae09","source":{"kind":"arxiv","id":"2308.14903","version":1},"attestation_state":"computed","paper":{"title":"MEMORY-VQ: Compression for Tractable Internet-Scale Memory","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Joshua Ainslie, Luke Vilnis, Michiel de Jong, Santiago Onta\\~n\\'on, Sumit Sanghai, William W. Cohen, Yury Zemlyanskiy","submitted_at":"2023-08-28T21:11:18Z","abstract_excerpt":"Retrieval augmentation is a powerful but expensive method to make language models more knowledgeable about the world. Memory-based methods like LUMEN pre-compute token representations for retrieved passages to drastically speed up inference. However, memory also leads to much greater storage requirements from storing pre-computed representations.\n  We propose MEMORY-VQ, a new method to reduce storage requirements of memory-augmented models without sacrificing performance. Our method uses a vector quantization variational autoencoder (VQ-VAE) to compress token representations. We apply MEMORY-V"},"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":"2308.14903","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-28T21:11:18Z","cross_cats_sorted":[],"title_canon_sha256":"7242c245770f4e1fd9057b08577fd764a1fea08374183adc1c3302f8832c1d61","abstract_canon_sha256":"72402883e0e001ad94b0050a6cf91441163e78257f20e75dbe45bb6d7ba86e20"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:45:40.945994Z","signature_b64":"rjQOX005n3AMOU7lYSRAyXPZOySHiRxXqEl7jEeMXEn1q2G4AjrMnZF2HcCndQmujSznxhTVW4rh1Zmn14XgDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ef3a0ace51706d262892842185f0dda8869435a05d74dcaa883b5e40364fae09","last_reissued_at":"2026-07-05T06:45:40.945630Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:45:40.945630Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MEMORY-VQ: Compression for Tractable Internet-Scale Memory","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Joshua Ainslie, Luke Vilnis, Michiel de Jong, Santiago Onta\\~n\\'on, Sumit Sanghai, William W. Cohen, Yury Zemlyanskiy","submitted_at":"2023-08-28T21:11:18Z","abstract_excerpt":"Retrieval augmentation is a powerful but expensive method to make language models more knowledgeable about the world. Memory-based methods like LUMEN pre-compute token representations for retrieved passages to drastically speed up inference. However, memory also leads to much greater storage requirements from storing pre-computed representations.\n  We propose MEMORY-VQ, a new method to reduce storage requirements of memory-augmented models without sacrificing performance. Our method uses a vector quantization variational autoencoder (VQ-VAE) to compress token representations. We apply MEMORY-V"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.14903","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2308.14903/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":"2308.14903","created_at":"2026-07-05T06:45:40.945682+00:00"},{"alias_kind":"arxiv_version","alias_value":"2308.14903v1","created_at":"2026-07-05T06:45:40.945682+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.14903","created_at":"2026-07-05T06:45:40.945682+00:00"},{"alias_kind":"pith_short_12","alias_value":"545AVTSROBWS","created_at":"2026-07-05T06:45:40.945682+00:00"},{"alias_kind":"pith_short_16","alias_value":"545AVTSROBWSMKES","created_at":"2026-07-05T06:45:40.945682+00:00"},{"alias_kind":"pith_short_8","alias_value":"545AVTSR","created_at":"2026-07-05T06:45:40.945682+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/545AVTSROBWSMKESQQQYL4G5VC","json":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC.json","graph_json":"https://pith.science/api/pith-number/545AVTSROBWSMKESQQQYL4G5VC/graph.json","events_json":"https://pith.science/api/pith-number/545AVTSROBWSMKESQQQYL4G5VC/events.json","paper":"https://pith.science/paper/545AVTSR"},"agent_actions":{"view_html":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC","download_json":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC.json","view_paper":"https://pith.science/paper/545AVTSR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2308.14903&json=true","fetch_graph":"https://pith.science/api/pith-number/545AVTSROBWSMKESQQQYL4G5VC/graph.json","fetch_events":"https://pith.science/api/pith-number/545AVTSROBWSMKESQQQYL4G5VC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC/action/storage_attestation","attest_author":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC/action/author_attestation","sign_citation":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC/action/citation_signature","submit_replication":"https://pith.science/pith/545AVTSROBWSMKESQQQYL4G5VC/action/replication_record"}},"created_at":"2026-07-05T06:45:40.945682+00:00","updated_at":"2026-07-05T06:45:40.945682+00:00"}