{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GRYQWRJ2732XXBIJ7EECZWLLYD","short_pith_number":"pith:GRYQWRJ2","schema_version":"1.0","canonical_sha256":"34710b453afef57b8509f9082cd96bc0ce5f1fdc6c065ee29248ff348b7d1a01","source":{"kind":"arxiv","id":"2603.19732","version":2},"attestation_state":"computed","paper":{"title":"Helix: A Dual-Helix Co-Evolutionary Multi-Agent System for Prompt Optimization and Question Reformulation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Helix jointly optimizes question reformulation and prompt instructions via a three-stage co-evolutionary multi-agent framework.","cross_cats":[],"primary_cat":"cs.MA","authors_text":"Kewen Zhu, Liping Yi, Qinghua Hu, Xiang Li, Zhiming Zhao","submitted_at":"2026-03-20T08:16:09Z","abstract_excerpt":"Automated prompt optimization (APO) aims to improve large language model performance by refining prompt instructions. However, existing methods are largely constrained by fixed prompt templates, limited search spaces, or single-sided optimization that treats user questions as immutable inputs. In practice, question formulation and prompt design are inherently interdependent: clearer question structures facilitate focused reasoning and task understanding, while effective prompts reveal better ways to organize and restate queries. Ignoring this coupling fundamentally limits the effectiveness and"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2603.19732","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.MA","submitted_at":"2026-03-20T08:16:09Z","cross_cats_sorted":[],"title_canon_sha256":"f18e46796a7013cfe138ebca1526dba118f935da4831ae74669fd97b5c03d9b6","abstract_canon_sha256":"49bd8160e3b3ae794198737db1f7bf408c6d55b38a26f4337920402706fe8b6d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:01:40.343860Z","signature_b64":"rt0GiNZDVIg05pJgQy2nMOOxJ+LravbHY1KGFctqzcX/QSr/JkTbVYrCm/+kIIOatVFTaiETXVsWaLmUYI64Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"34710b453afef57b8509f9082cd96bc0ce5f1fdc6c065ee29248ff348b7d1a01","last_reissued_at":"2026-05-20T00:01:40.343074Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:01:40.343074Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Helix: A Dual-Helix Co-Evolutionary Multi-Agent System for Prompt Optimization and Question Reformulation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Helix jointly optimizes question reformulation and prompt instructions via a three-stage co-evolutionary multi-agent framework.","cross_cats":[],"primary_cat":"cs.MA","authors_text":"Kewen Zhu, Liping Yi, Qinghua Hu, Xiang Li, Zhiming Zhao","submitted_at":"2026-03-20T08:16:09Z","abstract_excerpt":"Automated prompt optimization (APO) aims to improve large language model performance by refining prompt instructions. However, existing methods are largely constrained by fixed prompt templates, limited search spaces, or single-sided optimization that treats user questions as immutable inputs. In practice, question formulation and prompt design are inherently interdependent: clearer question structures facilitate focused reasoning and task understanding, while effective prompts reveal better ways to organize and restate queries. Ignoring this coupling fundamentally limits the effectiveness and"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We propose a unified multi-agent system (Helix) that jointly optimizes question reformulation and prompt instructions through a structured three-stage co-evolutionary framework... achieving up to 3.95% performance improvements across tasks with favorable optimization efficiency.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That question formulation and prompt design are inherently interdependent in a manner that allows dual-track co-evolution between specialized agents to produce complementary improvements unavailable to single-sided optimization methods.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Helix introduces a three-stage dual-helix co-evolutionary multi-agent framework that jointly optimizes question reformulation and prompt instructions, reporting up to 3.95% gains on 12 benchmarks versus 6 baselines.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Helix jointly optimizes question reformulation and prompt instructions via a three-stage co-evolutionary multi-agent framework.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"3785e8b305b4cb6878a8eea3e1d83681943589de06ea196e7fb0ab5a5216b2c1"},"source":{"id":"2603.19732","kind":"arxiv","version":2},"verdict":{"id":"aa2e58ce-8dd7-488e-9044-34df551825c9","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T18:23:40.956352Z","strongest_claim":"We propose a unified multi-agent system (Helix) that jointly optimizes question reformulation and prompt instructions through a structured three-stage co-evolutionary framework... achieving up to 3.95% performance improvements across tasks with favorable optimization efficiency.","one_line_summary":"Helix introduces a three-stage dual-helix co-evolutionary multi-agent framework that jointly optimizes question reformulation and prompt instructions, reporting up to 3.95% gains on 12 benchmarks versus 6 baselines.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That question formulation and prompt design are inherently interdependent in a manner that allows dual-track co-evolution between specialized agents to produce complementary improvements unavailable to single-sided optimization methods.","pith_extraction_headline":"Helix jointly optimizes question reformulation and prompt instructions via a three-stage co-evolutionary multi-agent framework."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.19732/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":34,"sample":[{"doi":"","year":null,"title":"Primary rule:Enclose key pronouns (e.g., [their]) in brackets within the provided sentence to ensure they are visually distinct and immediately identifiable","work_id":"77b6589d-5b31-4ae3-b86d-423b71f9551e","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Secondary rule:Apply the brackets uniformly to all instances of the pronoun in the sentence and related options to maintain consistency","work_id":"7bf2e012-e543-4e87-ae2c-38424c9c2503","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Table 11.Original question from the Disambiguation QA task before optimization","work_id":"78979823-670d-4344-ad8f-e8326f7e0d57","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Primary rule:Include a concise clarification in the question explicitly stating whether the path is open, closed, or subject to specific rules about near-collinear points, overlapping paths, or self-i","work_id":"c784fddc-aaf7-49a0-bef7-17bf7ea6c88f","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Secondary rule:Add this clarification as a note in parentheses or as a short sentence at the end of the question, ensuring it integrates naturally without disrupting readability or overloading the que","work_id":"9fb44aa6-65cc-4323-a04e-08bac851fd6c","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":34,"snapshot_sha256":"6685d17a29d41df487d9c90357670c6c08a7d70292188e24fb6c46e4285a7d9a","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"65a56f7d3bf7fb3d86a4050f37003fde65d59b9d2842db58e95cbeeeedfce3f3"},"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":"2603.19732","created_at":"2026-05-20T00:01:40.343183+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.19732v2","created_at":"2026-05-20T00:01:40.343183+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.19732","created_at":"2026-05-20T00:01:40.343183+00:00"},{"alias_kind":"pith_short_12","alias_value":"GRYQWRJ2732X","created_at":"2026-05-20T00:01:40.343183+00:00"},{"alias_kind":"pith_short_16","alias_value":"GRYQWRJ2732XXBIJ","created_at":"2026-05-20T00:01:40.343183+00:00"},{"alias_kind":"pith_short_8","alias_value":"GRYQWRJ2","created_at":"2026-05-20T00:01:40.343183+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.14585","citing_title":"Prompt Optimization Is a Coin Flip: Diagnosing When It Helps in Compound AI Systems","ref_index":15,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD","json":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD.json","graph_json":"https://pith.science/api/pith-number/GRYQWRJ2732XXBIJ7EECZWLLYD/graph.json","events_json":"https://pith.science/api/pith-number/GRYQWRJ2732XXBIJ7EECZWLLYD/events.json","paper":"https://pith.science/paper/GRYQWRJ2"},"agent_actions":{"view_html":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD","download_json":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD.json","view_paper":"https://pith.science/paper/GRYQWRJ2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.19732&json=true","fetch_graph":"https://pith.science/api/pith-number/GRYQWRJ2732XXBIJ7EECZWLLYD/graph.json","fetch_events":"https://pith.science/api/pith-number/GRYQWRJ2732XXBIJ7EECZWLLYD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD/action/storage_attestation","attest_author":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD/action/author_attestation","sign_citation":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD/action/citation_signature","submit_replication":"https://pith.science/pith/GRYQWRJ2732XXBIJ7EECZWLLYD/action/replication_record"}},"created_at":"2026-05-20T00:01:40.343183+00:00","updated_at":"2026-05-20T00:01:40.343183+00:00"}