基于账本控制的零样本自编排提升LLM编码性能研究
原标题：Zero-Shot Self-Orchestration with Ledger-Based Control for Improved LLM Coding Performance
AI 摘要
该研究比较了多智能体LLM系统与单模型基线在编码任务上的表现，发现引入管理器-工人脚手架的效果因模型而异：对部分模型（如Qwen3.8-27B、GPT-5.6-Luna等）有显著提升，对另一些模型（如Qwen3.6-35B）则无益或负面。管理器机制虽使token消耗增加约三倍，但相比升级更大模型，能以更低成本获得相近精度。研究还分析了增益机制，包括上下文管理和问题分解。
正文节选
Zero-Shot Self-Orchestration with Ledger-Based Control for Improved LLM Coding Performance Abstract Multi-agent large language model systems are widely reported to beat single-model baselines, but the evidence is mixed, and comparisons are usually confounded: pipelines change token budgets, tool calls, and prompts simultaneously, so an aggregate gain rarely reveals what actually helped. We investigate the effect of introducing the manager–worker scaffold over a shared filesystem workspace, with