{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ETOL6PWBPMYF25QP4IK7BDGLH5","short_pith_number":"pith:ETOL6PWB","schema_version":"1.0","canonical_sha256":"24dcbf3ec17b305d760fe215f08ccb3f571ff59eb7ec63e3cfcb3f94dbfe9374","source":{"kind":"arxiv","id":"2506.04185","version":1},"attestation_state":"computed","paper":{"title":"R-Search: Empowering LLM Reasoning with Search via Multi-Reward Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Daren Zha, Dingling Xu, Limin Liu, Qingfei Zhao, Ruobing Wang","submitted_at":"2025-06-04T17:29:22Z","abstract_excerpt":"Large language models (LLMs) have notably progressed in multi-step and long-chain reasoning. However, extending their reasoning capabilities to encompass deep interactions with search remains a non-trivial challenge, as models often fail to identify optimal reasoning-search interaction trajectories, resulting in suboptimal responses. We propose R-Search, a novel reinforcement learning framework for Reasoning-Search integration, designed to enable LLMs to autonomously execute multi-step reasoning with deep search interaction, and learn optimal reasoning search interaction trajectories via multi"},"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":"2506.04185","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-04T17:29:22Z","cross_cats_sorted":[],"title_canon_sha256":"bd3bec7636cf126b563a6226e2ec4fd55c345ac07d8e65f4e6c49c6a9c4a8eeb","abstract_canon_sha256":"264f7b115a358f53f7507348959998a77310c4b6ff588c7eddd4ae008ff2882d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:15:58.564640Z","signature_b64":"5M4NjzN+LCcGUE6jdOEkt8WalFqG7Ei+u+2aMKlneWJJmM3Fi0BlaX3fhIFTFHpswYTh7h4MsR6cV6w0sZDgCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"24dcbf3ec17b305d760fe215f08ccb3f571ff59eb7ec63e3cfcb3f94dbfe9374","last_reissued_at":"2026-07-05T11:15:58.564096Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:15:58.564096Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"R-Search: Empowering LLM Reasoning with Search via Multi-Reward Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Daren Zha, Dingling Xu, Limin Liu, Qingfei Zhao, Ruobing Wang","submitted_at":"2025-06-04T17:29:22Z","abstract_excerpt":"Large language models (LLMs) have notably progressed in multi-step and long-chain reasoning. However, extending their reasoning capabilities to encompass deep interactions with search remains a non-trivial challenge, as models often fail to identify optimal reasoning-search interaction trajectories, resulting in suboptimal responses. We propose R-Search, a novel reinforcement learning framework for Reasoning-Search integration, designed to enable LLMs to autonomously execute multi-step reasoning with deep search interaction, and learn optimal reasoning search interaction trajectories via multi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.04185","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/2506.04185/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":"2506.04185","created_at":"2026-07-05T11:15:58.564165+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.04185v1","created_at":"2026-07-05T11:15:58.564165+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.04185","created_at":"2026-07-05T11:15:58.564165+00:00"},{"alias_kind":"pith_short_12","alias_value":"ETOL6PWBPMYF","created_at":"2026-07-05T11:15:58.564165+00:00"},{"alias_kind":"pith_short_16","alias_value":"ETOL6PWBPMYF25QP","created_at":"2026-07-05T11:15:58.564165+00:00"},{"alias_kind":"pith_short_8","alias_value":"ETOL6PWB","created_at":"2026-07-05T11:15:58.564165+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21262","citing_title":"ARCO: Adaptive Rubrics with Co-Evolution for Multi-Step LLM-Based Agents","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27860","citing_title":"C-MIG: Multi-view Information Gain-based Retrieval-Augmented Generation for Clinical Diagnosis Reasoning","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26122","citing_title":"DocArena: Turning Raw Documents into Controllable Training Environments for Document Search Agents","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10923","citing_title":"Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2510.00861","citing_title":"Erase to Improve: Erasable Reinforcement Learning for Search-Augmented LLMs","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2510.11122","citing_title":"Learning to Trust: Dynamic Utilization of Retrieval-Augmented Generation for E-commerce Search Relevance","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10923","citing_title":"Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06285","citing_title":"LatentRAG: Latent Reasoning and Retrieval for Efficient Agentic RAG","ref_index":50,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5","json":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5.json","graph_json":"https://pith.science/api/pith-number/ETOL6PWBPMYF25QP4IK7BDGLH5/graph.json","events_json":"https://pith.science/api/pith-number/ETOL6PWBPMYF25QP4IK7BDGLH5/events.json","paper":"https://pith.science/paper/ETOL6PWB"},"agent_actions":{"view_html":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5","download_json":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5.json","view_paper":"https://pith.science/paper/ETOL6PWB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.04185&json=true","fetch_graph":"https://pith.science/api/pith-number/ETOL6PWBPMYF25QP4IK7BDGLH5/graph.json","fetch_events":"https://pith.science/api/pith-number/ETOL6PWBPMYF25QP4IK7BDGLH5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5/action/storage_attestation","attest_author":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5/action/author_attestation","sign_citation":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5/action/citation_signature","submit_replication":"https://pith.science/pith/ETOL6PWBPMYF25QP4IK7BDGLH5/action/replication_record"}},"created_at":"2026-07-05T11:15:58.564165+00:00","updated_at":"2026-07-05T11:15:58.564165+00:00"}