{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:5WWTDNK72RYSVRRQ5U4QLWVRAG","short_pith_number":"pith:5WWTDNK7","schema_version":"1.0","canonical_sha256":"edad31b55fd4712ac630ed3905dab10188c2e1e9f828a4f45a77a9dbfcbdd1b4","source":{"kind":"arxiv","id":"2405.12363","version":2},"attestation_state":"computed","paper":{"title":"Question-Based Retrieval using Atomic Units for Enterprise RAG","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Mark Gales, Vatsal Raina","submitted_at":"2024-05-20T20:27:00Z","abstract_excerpt":"Enterprise retrieval augmented generation (RAG) offers a highly flexible framework for combining powerful large language models (LLMs) with internal, possibly temporally changing, documents. In RAG, documents are first chunked. Relevant chunks are then retrieved for a user query, which are passed as context to a synthesizer LLM to generate the query response. However, the retrieval step can limit performance, as incorrect chunks can lead the synthesizer LLM to generate a false response. This work applies a zero-shot adaptation of standard dense retrieval steps for more accurate chunk recall. S"},"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":"2405.12363","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-05-20T20:27:00Z","cross_cats_sorted":[],"title_canon_sha256":"722bacb50f0885181669458dcecd407b7cd7d18990720182c6366aa14ece9972","abstract_canon_sha256":"fd3830f2094f97ceafcf8bf5d21675cc2a349590abf9bff6d2020c706f9c1e05"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:01:00.352861Z","signature_b64":"VxofetNTz9V0AKubTAIU4N8hn8+1/ppQpGiMmFHw/ZdPXvN2R5vQ7RFiXIGRDt1q//glAB5o+J5KL/UxTEDXBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"edad31b55fd4712ac630ed3905dab10188c2e1e9f828a4f45a77a9dbfcbdd1b4","last_reissued_at":"2026-07-05T09:01:00.352431Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:01:00.352431Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Question-Based Retrieval using Atomic Units for Enterprise RAG","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Mark Gales, Vatsal Raina","submitted_at":"2024-05-20T20:27:00Z","abstract_excerpt":"Enterprise retrieval augmented generation (RAG) offers a highly flexible framework for combining powerful large language models (LLMs) with internal, possibly temporally changing, documents. In RAG, documents are first chunked. Relevant chunks are then retrieved for a user query, which are passed as context to a synthesizer LLM to generate the query response. However, the retrieval step can limit performance, as incorrect chunks can lead the synthesizer LLM to generate a false response. This work applies a zero-shot adaptation of standard dense retrieval steps for more accurate chunk recall. S"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.12363","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/2405.12363/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":"2405.12363","created_at":"2026-07-05T09:01:00.352487+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.12363v2","created_at":"2026-07-05T09:01:00.352487+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.12363","created_at":"2026-07-05T09:01:00.352487+00:00"},{"alias_kind":"pith_short_12","alias_value":"5WWTDNK72RYS","created_at":"2026-07-05T09:01:00.352487+00:00"},{"alias_kind":"pith_short_16","alias_value":"5WWTDNK72RYSVRRQ","created_at":"2026-07-05T09:01:00.352487+00:00"},{"alias_kind":"pith_short_8","alias_value":"5WWTDNK7","created_at":"2026-07-05T09:01:00.352487+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2402.19473","citing_title":"Retrieval-Augmented Generation for AI-Generated Content: A Survey","ref_index":138,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG","json":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG.json","graph_json":"https://pith.science/api/pith-number/5WWTDNK72RYSVRRQ5U4QLWVRAG/graph.json","events_json":"https://pith.science/api/pith-number/5WWTDNK72RYSVRRQ5U4QLWVRAG/events.json","paper":"https://pith.science/paper/5WWTDNK7"},"agent_actions":{"view_html":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG","download_json":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG.json","view_paper":"https://pith.science/paper/5WWTDNK7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.12363&json=true","fetch_graph":"https://pith.science/api/pith-number/5WWTDNK72RYSVRRQ5U4QLWVRAG/graph.json","fetch_events":"https://pith.science/api/pith-number/5WWTDNK72RYSVRRQ5U4QLWVRAG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG/action/storage_attestation","attest_author":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG/action/author_attestation","sign_citation":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG/action/citation_signature","submit_replication":"https://pith.science/pith/5WWTDNK72RYSVRRQ5U4QLWVRAG/action/replication_record"}},"created_at":"2026-07-05T09:01:00.352487+00:00","updated_at":"2026-07-05T09:01:00.352487+00:00"}