{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:Y3FVWKAY3MJARCIYNOTBAXPM2D","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"15ec28207e7fa01c11f4ba4113c5404c7374bdf2c3e3ae988b5c9ae3df33cc15","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T13:46:35Z","title_canon_sha256":"5ce489a8eca4b57f6e76fbf9df1ffef77b147666f4ede7b2acb0996faabc7c97"},"schema_version":"1.0","source":{"id":"2605.13534","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13534","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13534v1","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13534","created_at":"2026-05-18T02:44:24Z"},{"alias_kind":"pith_short_12","alias_value":"Y3FVWKAY3MJA","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"Y3FVWKAY3MJARCIY","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"Y3FVWKAY","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:a9e68b5d40052b340e32bb04b76cc76faf5f8ffc6e5af1719bf71949de65f639","target":"graph","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"MultiSearch outperforms baseline methods, enhancing the SNR of retrieval and improving reasoning performance in question-answering tasks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That generating multiple queries from different perspectives and then performing explicit merging will consistently raise signal-to-noise ratio without introducing new noise sources or prohibitive compute overhead in the merging step."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"MultiSearch uses parallel multi-query retrieval plus explicit merging inside a reinforcement-learning loop to improve retrieval-augmented reasoning, outperforming baselines on seven QA benchmarks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"MultiSearch retrieves external knowledge from multiple query perspectives in parallel and merges the results to raise signal-to-noise ratio before reasoning."}],"snapshot_sha256":"4447f975051f44b691ac6509fb9d35f265808db1e87e7a8011424fe6c8bd4778"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"1d77c544fa3c3b297707671fad1749160766881d74a58c621c4146a3e839d2ef"},"paper":{"abstract_excerpt":"Deep search agents have proven effective in enhancing LLMs by retrieving external knowledge during multi-step reasoning. However, existing methods often generate a single query for retrieval at each reasoning step, limiting information coverage and introducing high noise. This may result in low signal-to-noise ratios (SNR) during search, degrading reasoning accuracy and leading to unnecessary reasoning steps. In this paper, we introduce MultiSearch, an RL-based framework that addresses these limitations through multi-query retrieval and explicit merging of retrieved information. At each reason","authors_text":"Chunxu Shen, Jiabei Liu, Jiancan Wu, Junfei Tan, Lingling Yi, Wenyu Mao, Xiang Wang","cross_cats":[],"headline":"MultiSearch retrieves external knowledge from multiple query perspectives in parallel and merges the results to raise signal-to-noise ratio before reasoning.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T13:46:35Z","title":"Scaling Retrieval-Augmented Reasoning with Parallel Search and Explicit Merging"},"references":{"count":47,"internal_anchors":13,"resolved_work":47,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"A survey on evaluation of large language models.ACM transactions on intelligent systems and technology, 15(3):1–45","work_id":"a9075135-e1ef-4f62-89ff-20df37c58319","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Siren’s song in the ai ocean: A survey on hallucination in large language models.Computational Linguistics, 51(4):1373–1418, 2025","work_id":"0ddc813d-938a-47c5-b4a0-693d9a296cfd","year":2025},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"A study of generative large language model for medical research and healthcare.NPJ digital medicine, 6(1):210, 2023","work_id":"06400e59-275e-40c4-9fa0-50ae5f9f4128","year":2023},{"cited_arxiv_id":"2312.10997","doi":"","is_internal_anchor":true,"ref_index":4,"title":"Retrieval-Augmented Generation for Large Language Models: A Survey","work_id":"b80d2790-6cd9-4c87-b3c4-de404f99a80e","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Retrieval-augmented generation for knowledge-intensive nlp tasks.Advances in neural information processing systems, 33:9459–9474","work_id":"f4e7687c-6c0e-4b2d-92f6-01888410340c","year":2020}],"snapshot_sha256":"09bd96588df310aad790ab67f530742764c40f0936eb1c84757f7770a33c50dc"},"source":{"id":"2605.13534","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T18:49:24.839275Z","id":"cfb4a029-898a-4fcd-b535-0ff45b2448f8","model_set":{"reader":"grok-4.3"},"one_line_summary":"MultiSearch uses parallel multi-query retrieval plus explicit merging inside a reinforcement-learning loop to improve retrieval-augmented reasoning, outperforming baselines on seven QA benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"MultiSearch retrieves external knowledge from multiple query perspectives in parallel and merges the results to raise signal-to-noise ratio before reasoning.","strongest_claim":"MultiSearch outperforms baseline methods, enhancing the SNR of retrieval and improving reasoning performance in question-answering tasks.","weakest_assumption":"That generating multiple queries from different perspectives and then performing explicit merging will consistently raise signal-to-noise ratio without introducing new noise sources or prohibitive compute overhead in the merging step."}},"verdict_id":"cfb4a029-898a-4fcd-b535-0ff45b2448f8"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5f6e1ebb7047085c31fa0259044a59b15d3d1624203b4bd6de743bd102179813","target":"record","created_at":"2026-05-18T02:44:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"15ec28207e7fa01c11f4ba4113c5404c7374bdf2c3e3ae988b5c9ae3df33cc15","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-05-13T13:46:35Z","title_canon_sha256":"5ce489a8eca4b57f6e76fbf9df1ffef77b147666f4ede7b2acb0996faabc7c97"},"schema_version":"1.0","source":{"id":"2605.13534","kind":"arxiv","version":1}},"canonical_sha256":"c6cb5b2818db120889186ba6105decd0f6dc856362d7ed29b403d7ea21b893b2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c6cb5b2818db120889186ba6105decd0f6dc856362d7ed29b403d7ea21b893b2","first_computed_at":"2026-05-18T02:44:24.185809Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:24.185809Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"behxxQwQslhz2+tTHeS95ZFvOkva1W+F93JA5i1pmSwR/EPdWLVVJ8FV8U6dEaU1Rdngh6rV+/PF/59j+QLdAQ==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:24.186888Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13534","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5f6e1ebb7047085c31fa0259044a59b15d3d1624203b4bd6de743bd102179813","sha256:a9e68b5d40052b340e32bb04b76cc76faf5f8ffc6e5af1719bf71949de65f639"],"state_sha256":"bbbe049c0c484e1cdc1826b49c3bf066f82189e79610682f7491aa52d558b308"}