{"ID":2853497,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.18893","arxiv_id":"2510.18893","title":"CodeCRDT: Observation-Driven Coordination for Multi-Agent LLM Code Generation","abstract":"Multi-agent LLM systems fail to realize parallel speedups due to costly coordination. We present CodeCRDT, an observation-driven coordination pattern where agents coordinate by monitoring a shared state with observable updates and deterministic convergence, rather than explicit message passing. Using Conflict-Free Replicated Data Types (CRDTs), CodeCRDT enables lock-free, conflict-free concurrent code generation with strong eventual consistency. Evaluation across 600 trials (6 tasks, 50 runs per mode) shows both benefits and trade-offs: up to 21.1% speedup on some tasks, up to 39.4% slowdown on others, and 100% convergence with zero merge failures. The study formalizes observation-driven coordination for stochastic LLM agents, revealing semantic conflict rates (5-10%) and quality-performance tradeoffs, and provides empirical characterization of when parallel coordination succeeds versus fails based on task structure.","short_abstract":"Multi-agent LLM systems fail to realize parallel speedups due to costly coordination. We present CodeCRDT, an observation-driven coordination pattern where agents coordinate by monitoring a shared state with observable updates and deterministic convergence, rather than explicit message passing. Using Conflict-Free Repl...","url_abs":"https://arxiv.org/abs/2510.18893","url_pdf":"https://arxiv.org/pdf/2510.18893v1","authors":"[\"Sergey Pugachev\"]","published":"2025-10-18T20:50:01Z","proceeding":"cs.DC","tasks":"[\"cs.DC\",\"cs.AI\",\"cs.SE\"]","methods":"[\"Large Language Model\"]","has_code":false}