{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OYBNBUGQ5UUIWPNTGQRAOYDL6S","short_pith_number":"pith:OYBNBUGQ","schema_version":"1.0","canonical_sha256":"7602d0d0d0ed288b3db3342207606bf48e6b111580f5b046cda04b248cdcdfc2","source":{"kind":"arxiv","id":"2508.19113","version":3},"attestation_state":"computed","paper":{"title":"Hybrid Deep Searcher: Scalable Parallel and Sequential Search Reasoning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Dahyun Lee, Dayoon Ko, Gunhee Kim, Haeju Park, Jihyuk Kim, Kyungjae Lee, Moontae Lee, Sohyeon Kim, Yongrae Jo","submitted_at":"2025-08-26T15:15:17Z","abstract_excerpt":"Large reasoning models (LRMs) combined with retrieval-augmented generation (RAG) have enabled deep research agents capable of multi-step reasoning with external knowledge retrieval. However, we find that existing approaches rarely demonstrate test-time search scaling. Methods that extend reasoning through single-query sequential search suffer from limited evidence coverage, while approaches that generate multiple independent queries per step often lack structured aggregation, hindering deeper sequential reasoning. We propose a hybrid search strategy to address these limitations. We introduce H"},"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":"2508.19113","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.AI","submitted_at":"2025-08-26T15:15:17Z","cross_cats_sorted":[],"title_canon_sha256":"3f9909f226b010ba18d8b1ed93c426653f890e054a7168a231a0a70c404b71af","abstract_canon_sha256":"1459d05d22a3ddfdb73d8380b3accb54fde0fed96759ebf286d52637d1b6fac6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:05:01.614758Z","signature_b64":"vACs4plZFIgapxMBEm08NMcdjoAc/kzu1xBRgVuVkuyWNxmz9hjtwakYSbvQv8pUD742wAfPkMfOXTNr4+ikBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7602d0d0d0ed288b3db3342207606bf48e6b111580f5b046cda04b248cdcdfc2","last_reissued_at":"2026-05-26T02:05:01.613892Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:05:01.613892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Hybrid Deep Searcher: Scalable Parallel and Sequential Search Reasoning","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Dahyun Lee, Dayoon Ko, Gunhee Kim, Haeju Park, Jihyuk Kim, Kyungjae Lee, Moontae Lee, Sohyeon Kim, Yongrae Jo","submitted_at":"2025-08-26T15:15:17Z","abstract_excerpt":"Large reasoning models (LRMs) combined with retrieval-augmented generation (RAG) have enabled deep research agents capable of multi-step reasoning with external knowledge retrieval. However, we find that existing approaches rarely demonstrate test-time search scaling. Methods that extend reasoning through single-query sequential search suffer from limited evidence coverage, while approaches that generate multiple independent queries per step often lack structured aggregation, hindering deeper sequential reasoning. We propose a hybrid search strategy to address these limitations. We introduce H"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.19113","kind":"arxiv","version":3},"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/2508.19113/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":"2508.19113","created_at":"2026-05-26T02:05:01.614008+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.19113v3","created_at":"2026-05-26T02:05:01.614008+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.19113","created_at":"2026-05-26T02:05:01.614008+00:00"},{"alias_kind":"pith_short_12","alias_value":"OYBNBUGQ5UUI","created_at":"2026-05-26T02:05:01.614008+00:00"},{"alias_kind":"pith_short_16","alias_value":"OYBNBUGQ5UUIWPNT","created_at":"2026-05-26T02:05:01.614008+00:00"},{"alias_kind":"pith_short_8","alias_value":"OYBNBUGQ","created_at":"2026-05-26T02:05:01.614008+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2510.00861","citing_title":"Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMs","ref_index":48,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S","json":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S.json","graph_json":"https://pith.science/api/pith-number/OYBNBUGQ5UUIWPNTGQRAOYDL6S/graph.json","events_json":"https://pith.science/api/pith-number/OYBNBUGQ5UUIWPNTGQRAOYDL6S/events.json","paper":"https://pith.science/paper/OYBNBUGQ"},"agent_actions":{"view_html":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S","download_json":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S.json","view_paper":"https://pith.science/paper/OYBNBUGQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.19113&json=true","fetch_graph":"https://pith.science/api/pith-number/OYBNBUGQ5UUIWPNTGQRAOYDL6S/graph.json","fetch_events":"https://pith.science/api/pith-number/OYBNBUGQ5UUIWPNTGQRAOYDL6S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S/action/storage_attestation","attest_author":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S/action/author_attestation","sign_citation":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S/action/citation_signature","submit_replication":"https://pith.science/pith/OYBNBUGQ5UUIWPNTGQRAOYDL6S/action/replication_record"}},"created_at":"2026-05-26T02:05:01.614008+00:00","updated_at":"2026-05-26T02:05:01.614008+00:00"}