{"paper":{"title":"Co-evolving Agent Architectures and Interpretable Reasoning for Automated Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Representing agent workflows as evolvable AOE-style networks and co-evolving their topologies with reasoning paths improves automated operations research performance and adds structural interpretability.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Jiahao Huang, Peilan Xu, Wenjian Luo, Xiaoya Nan","submitted_at":"2026-04-20T01:44:18Z","abstract_excerpt":"Automating operations research (OR) with large language models (LLMs) remains limited by hand-crafted reasoning--execution workflows. Complex OR tasks require adaptive coordination among problem interpretation, mathematical formulation, solver selection, code generation, and iterative debugging. To address this limitation, we propose EvoOR-Agent, a co-evolutionary framework for automated optimization. The framework represents agent workflows as activity-on-edge (AOE)-style networks, making workflow topology, execution dependencies, and alternative reasoning paths explicit. On this representati"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Empirical results on heterogeneous OR benchmarks show that the proposed framework consistently improves over zero-shot LLMs, fixed-pipeline OR agents, and representative evolutionary agent frameworks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That representing agent workflows as AOE-style networks and applying graph-mediated path-conditioned recombination plus multi-granularity semantic mutation will produce meaningful, generalizable improvements in complex OR reasoning and execution.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"EvoOR-Agent co-evolves agent architectures as AOE-style networks with graph-mediated recombination and knowledge-base-assisted mutation to outperform fixed LLM pipelines on OR benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Representing agent workflows as evolvable AOE-style networks and co-evolving their topologies with reasoning paths improves automated operations research performance and adds structural interpretability.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"a6ada9add0211e2fa4b3939ffb0ce442019d772f7b30f4e01c4043b0516831fd"},"source":{"id":"2604.17708","kind":"arxiv","version":2},"verdict":{"id":"473cef7d-6563-44da-b1d0-8fb64f7515b0","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T05:18:08.819895Z","strongest_claim":"Empirical results on heterogeneous OR benchmarks show that the proposed framework consistently improves over zero-shot LLMs, fixed-pipeline OR agents, and representative evolutionary agent frameworks.","one_line_summary":"EvoOR-Agent co-evolves agent architectures as AOE-style networks with graph-mediated recombination and knowledge-base-assisted mutation to outperform fixed LLM pipelines on OR benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That representing agent workflows as AOE-style networks and applying graph-mediated path-conditioned recombination plus multi-granularity semantic mutation will produce meaningful, generalizable improvements in complex OR reasoning and execution.","pith_extraction_headline":"Representing agent workflows as evolvable AOE-style networks and co-evolving their topologies with reasoning paths improves automated operations research performance and adds structural interpretability."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.17708/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"}