{"paper":{"title":"Dual-Cluster Memory Agent: Resolving Multi-Paradigm Ambiguity in Optimization Problem Solving","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Dual clusters of historical solutions let LLMs resolve conflicting modeling paradigms in optimization problems.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bifan Wei, Boxuan Zhang, Jun Liu, Lingling Zhang, Xinyu Zhang, Yuchen Wan, Zesheng Yang","submitted_at":"2026-04-22T04:55:31Z","abstract_excerpt":"Large Language Models (LLMs) often struggle with structural ambiguity in optimization problems, where a single problem admits multiple related but conflicting modeling paradigms, hindering effective solution generation. To address this, we propose Dual-Cluster Memory Agent (DCM-Agent) to enhance performance by leveraging historical solutions in a training-free manner. Central to this is Dual-Cluster Memory Construction. This agent assigns historical solutions to modeling and coding clusters, then distills each cluster's content into three structured types: Approach, Checklist, and Pitfall. Thi"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"DCM-Agent achieves an average performance improvement of 11%-21% across seven optimization benchmarks, with a knowledge inheritance phenomenon where memory constructed by larger models guides smaller models toward superior performance.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That historical solutions from prior problems can be reliably clustered and distilled into generalizable, non-overfitting guidance that transfers to new optimization instances with multi-paradigm ambiguity.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"DCM-Agent improves LLM performance on multi-paradigm optimization problems by 11-21% via dual-cluster memory construction and dynamic inference guidance.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Dual clusters of historical solutions let LLMs resolve conflicting modeling paradigms in optimization problems.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7e8154bdcaecc49918e6089df59cfb30bd76fb6402a863bc2f29b706664565b1"},"source":{"id":"2604.20183","kind":"arxiv","version":2},"verdict":{"id":"9a3a5405-9877-4ed3-a1de-f8d133860cbe","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T00:35:26.824440Z","strongest_claim":"DCM-Agent achieves an average performance improvement of 11%-21% across seven optimization benchmarks, with a knowledge inheritance phenomenon where memory constructed by larger models guides smaller models toward superior performance.","one_line_summary":"DCM-Agent improves LLM performance on multi-paradigm optimization problems by 11-21% via dual-cluster memory construction and dynamic inference guidance.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That historical solutions from prior problems can be reliably clustered and distilled into generalizable, non-overfitting guidance that transfers to new optimization instances with multi-paradigm ambiguity.","pith_extraction_headline":"Dual clusters of historical solutions let LLMs resolve conflicting modeling paradigms in optimization problems."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.20183/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T15:34:58.751657Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-20T02:14:30.263284Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"869f8ff737f3b0d89b9ad828ce5008896efc343a5d6036473ac7689749e6aad5"},"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"}