用文本梯度优化自动生成可复现的仿真器代码。
SOCIA-Nabla: Textual Gradient Meets Multi-Agent Orchestration for Automated Simulator Generation
- 将仿真器构建视为代码在文本计算图中的优化问题。
- 在三个真实任务中达到当前最佳准确率，且无需专家深度干预。
- 适合需要自动化仿真开发的跨领域研究者与工程团队。
本文提出SOCIA-Nabla，一个端到端的智能体框架，将仿真器构建视为在文本计算图中对代码进行实例优化的过程。专用LLM驱动的智能体作为图节点，工作流管理器执行损失驱动循环：代码生成 -> 执行 -> 评估 -> 修复。优化器采用文本梯度下降（TGD），人类仅在任务确认时介入，极大降低专家负担，同时以代码本身为可训练对象。在用户建模、遮罩采用和个性化出行三个物理信息系统任务中，SOCIA-Nabla均达到当前最优整体准确率。通过融合多智能体编排与损失对齐优化视角，该框架将脆弱的提示流水线转化为可复现、约束感知的仿真代码生成系统，支持跨领域与不同粒度的扩展。本工作正在审稿中，代码即将开源。
原文摘要 · Abstract (English)
In this paper, we present SOCIA-Nabla, an end-to-end, agentic framework that treats simulator construction asinstance optimization over code within a textual computation graph. Specialized LLM-driven agents are embedded as graph nodes, and a workflow manager executes a loss-driven loop: code synthesis -> execution -> evaluation -> code repair. The optimizer performs Textual-Gradient Descent (TGD), while human-in-the-loop interaction is reserved for task-spec confirmation, minimizing expert effort and keeping the code itself as the trainable object. Across three CPS tasks, i.e., User Modeling, Mask Adoption, and Personal Mobility, SOCIA-Nabla attains state-of-the-art overall accuracy. By unifying multi-agent orchestration with a loss-aligned optimization view, SOCIA-Nabla converts brittle prompt pipelines into reproducible, constraint-aware simulator code generation that scales across domains and simulation granularities. This work is under review, and we will release the code soon.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。