{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HFX7EDCGU7KMUL4U6FDHVSP2UT","short_pith_number":"pith:HFX7EDCG","schema_version":"1.0","canonical_sha256":"396ff20c46a7d4ca2f94f1467ac9faa4c74aa1abfb07e61d09d3189d18071abc","source":{"kind":"arxiv","id":"2401.14021","version":1},"attestation_state":"computed","paper":{"title":"Accelerating Retrieval-Augmented Language Model Serving with Speculation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.LG","authors_text":"Alan Zhu, Lanting Li, Lijie Yang, Phitchaya Mangpo Phothilimthana, Yihua Xu, Zhihao Jia, Zhihao Zhang","submitted_at":"2024-01-25T09:06:44Z","abstract_excerpt":"Retrieval-augmented language models (RaLM) have demonstrated the potential to solve knowledge-intensive natural language processing (NLP) tasks by combining a non-parametric knowledge base with a parametric language model. Instead of fine-tuning a fully parametric model, RaLM excels at its low-cost adaptation to the latest data and better source attribution mechanisms. Among various RaLM approaches, iterative RaLM delivers a better generation quality due to a more frequent interaction between the retriever and the language model. Despite the benefits, iterative RaLM usually encounters high ove"},"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":"2401.14021","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-25T09:06:44Z","cross_cats_sorted":["cs.CL","cs.IR"],"title_canon_sha256":"44d63c3fac9b3a87c288398794e9507967a2edcd8a0e81599c1e4b96074658d6","abstract_canon_sha256":"88136dc2c846123308e1837646fe607ce27bd257bfa591661f3c6fc35c69951c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:37:21.061855Z","signature_b64":"GqaMeujBBjmefg1kYA6lT3E/2ndjgUrgRVY8AmVQFGToZ8N8obX51kd13di+rzEb3/6WGhsbZ+DyQWRVS8ktBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"396ff20c46a7d4ca2f94f1467ac9faa4c74aa1abfb07e61d09d3189d18071abc","last_reissued_at":"2026-07-05T07:37:21.061272Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:37:21.061272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Accelerating Retrieval-Augmented Language Model Serving with Speculation","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.LG","authors_text":"Alan Zhu, Lanting Li, Lijie Yang, Phitchaya Mangpo Phothilimthana, Yihua Xu, Zhihao Jia, Zhihao Zhang","submitted_at":"2024-01-25T09:06:44Z","abstract_excerpt":"Retrieval-augmented language models (RaLM) have demonstrated the potential to solve knowledge-intensive natural language processing (NLP) tasks by combining a non-parametric knowledge base with a parametric language model. Instead of fine-tuning a fully parametric model, RaLM excels at its low-cost adaptation to the latest data and better source attribution mechanisms. Among various RaLM approaches, iterative RaLM delivers a better generation quality due to a more frequent interaction between the retriever and the language model. Despite the benefits, iterative RaLM usually encounters high ove"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.14021","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/2401.14021/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":"2401.14021","created_at":"2026-07-05T07:37:21.061342+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.14021v1","created_at":"2026-07-05T07:37:21.061342+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.14021","created_at":"2026-07-05T07:37:21.061342+00:00"},{"alias_kind":"pith_short_12","alias_value":"HFX7EDCGU7KM","created_at":"2026-07-05T07:37:21.061342+00:00"},{"alias_kind":"pith_short_16","alias_value":"HFX7EDCGU7KMUL4U","created_at":"2026-07-05T07:37:21.061342+00:00"},{"alias_kind":"pith_short_8","alias_value":"HFX7EDCG","created_at":"2026-07-05T07:37:21.061342+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20948","citing_title":"Memory Grafting: Scaling Language Model Pre-training via Offline Conditional Memory","ref_index":57,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT","json":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT.json","graph_json":"https://pith.science/api/pith-number/HFX7EDCGU7KMUL4U6FDHVSP2UT/graph.json","events_json":"https://pith.science/api/pith-number/HFX7EDCGU7KMUL4U6FDHVSP2UT/events.json","paper":"https://pith.science/paper/HFX7EDCG"},"agent_actions":{"view_html":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT","download_json":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT.json","view_paper":"https://pith.science/paper/HFX7EDCG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.14021&json=true","fetch_graph":"https://pith.science/api/pith-number/HFX7EDCGU7KMUL4U6FDHVSP2UT/graph.json","fetch_events":"https://pith.science/api/pith-number/HFX7EDCGU7KMUL4U6FDHVSP2UT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT/action/storage_attestation","attest_author":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT/action/author_attestation","sign_citation":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT/action/citation_signature","submit_replication":"https://pith.science/pith/HFX7EDCGU7KMUL4U6FDHVSP2UT/action/replication_record"}},"created_at":"2026-07-05T07:37:21.061342+00:00","updated_at":"2026-07-05T07:37:21.061342+00:00"}