{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TF2H42PGTNLXBK7W3ZNJ2LOVDS","short_pith_number":"pith:TF2H42PG","schema_version":"1.0","canonical_sha256":"99747e69e69b5770abf6de5a9d2dd51c957d1d373917a537c7f28c5cb21236e9","source":{"kind":"arxiv","id":"2502.20330","version":2},"attestation_state":"computed","paper":{"title":"RAPID: Long-Context Inference with Retrieval-Augmented Speculative Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guanzheng Chen, Jinjie Ni, Michael Qizhe Shieh, Qilong Feng, Xin Li","submitted_at":"2025-02-27T17:59:36Z","abstract_excerpt":"The emergence of long-context large language models (LLMs) offers a promising alternative to traditional retrieval-augmented generation (RAG) for processing extensive documents. However, the computational overhead of long-context inference presents significant efficiency challenges. While Speculative Decoding (SD) traditionally accelerates inference using smaller draft models, its effectiveness diminishes substantially in long-context scenarios due to memory-bound KV cache operations. We introduce Retrieval-Augmented Speculative Decoding (RAPID), which leverages RAG for both accelerating and 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":"2502.20330","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-27T17:59:36Z","cross_cats_sorted":[],"title_canon_sha256":"72bfb3ab87790791437a38a4dbb2121b8bde7e3af9e90198b64a8f0c8f1b7fec","abstract_canon_sha256":"a8e4418943e0a1cc56cd7f7f43b0a4c4459960775a5914c0b02829877f242b07"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:25:42.454341Z","signature_b64":"vwk3R02RX8Iu9yNqDDuahtR37wqQf9WBtkvkLGpejBWvk3KOUjO/qr4PGyTrugbZQ+F3oG+RKoy8VqPGfezjAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"99747e69e69b5770abf6de5a9d2dd51c957d1d373917a537c7f28c5cb21236e9","last_reissued_at":"2026-07-05T11:25:42.453923Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:25:42.453923Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RAPID: Long-Context Inference with Retrieval-Augmented Speculative Decoding","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guanzheng Chen, Jinjie Ni, Michael Qizhe Shieh, Qilong Feng, Xin Li","submitted_at":"2025-02-27T17:59:36Z","abstract_excerpt":"The emergence of long-context large language models (LLMs) offers a promising alternative to traditional retrieval-augmented generation (RAG) for processing extensive documents. However, the computational overhead of long-context inference presents significant efficiency challenges. While Speculative Decoding (SD) traditionally accelerates inference using smaller draft models, its effectiveness diminishes substantially in long-context scenarios due to memory-bound KV cache operations. We introduce Retrieval-Augmented Speculative Decoding (RAPID), which leverages RAG for both accelerating and e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.20330","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/2502.20330/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":"2502.20330","created_at":"2026-07-05T11:25:42.453979+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.20330v2","created_at":"2026-07-05T11:25:42.453979+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.20330","created_at":"2026-07-05T11:25:42.453979+00:00"},{"alias_kind":"pith_short_12","alias_value":"TF2H42PGTNLX","created_at":"2026-07-05T11:25:42.453979+00:00"},{"alias_kind":"pith_short_16","alias_value":"TF2H42PGTNLXBK7W","created_at":"2026-07-05T11:25:42.453979+00:00"},{"alias_kind":"pith_short_8","alias_value":"TF2H42PG","created_at":"2026-07-05T11:25:42.453979+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20104","citing_title":"Draft Less, Retrieve More: Hybrid Tree Construction for Speculative Decoding","ref_index":5,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS","json":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS.json","graph_json":"https://pith.science/api/pith-number/TF2H42PGTNLXBK7W3ZNJ2LOVDS/graph.json","events_json":"https://pith.science/api/pith-number/TF2H42PGTNLXBK7W3ZNJ2LOVDS/events.json","paper":"https://pith.science/paper/TF2H42PG"},"agent_actions":{"view_html":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS","download_json":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS.json","view_paper":"https://pith.science/paper/TF2H42PG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.20330&json=true","fetch_graph":"https://pith.science/api/pith-number/TF2H42PGTNLXBK7W3ZNJ2LOVDS/graph.json","fetch_events":"https://pith.science/api/pith-number/TF2H42PGTNLXBK7W3ZNJ2LOVDS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS/action/storage_attestation","attest_author":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS/action/author_attestation","sign_citation":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS/action/citation_signature","submit_replication":"https://pith.science/pith/TF2H42PGTNLXBK7W3ZNJ2LOVDS/action/replication_record"}},"created_at":"2026-07-05T11:25:42.453979+00:00","updated_at":"2026-07-05T11:25:42.453979+00:00"}