MOSAIC用多智能体协作自动生成严谨的科学代码。
MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific Coding
- 设计专用智能体在师生框架中自我反思、推理、编码与调试。
- 在科学编码基准上准确率显著优于现有方法。
- 适合需要复杂推理和迭代的科研代码生成任务。
我们提出MOSAIC，一种用于解决复杂科学编程任务的多智能体大语言模型框架。与通用编程不同，科学工作流要求算法严谨、深度结合领域知识、具备领域特定推理能力，并支持无需输入输出测试用例的算法迭代。许多科学问题需分步求解一系列子问题以达成最终结果。MOSAIC为无训练框架，通过特化智能体在学生-教师范式下实现自我反思、构建推理过程、编写代码并进行调试，有效应对科学代码生成挑战。该设计支持逐步问题分解、精准错误修正，并结合我们提出的整合上下文窗口（CCW），在涉及链式子问题的复杂任务中缓解了大模型幻觉问题。我们在科学编程基准上评估MOSAIC，结果表明其在准确性、鲁棒性和可解释性方面均优于现有方法。
原文摘要 · Abstract (English)
We present MOSAIC, a multi-agent Large Language Model (LLM) framework for solving challenging scientific coding tasks. Unlike general-purpose coding, scientific workflows require algorithms that are rigorous, interconnected with deep domain knowledge, and incorporate domain-specific reasoning, as well as algorithm iteration without requiring I/O test cases. Many scientific problems also require a sequence of subproblems to be solved, leading to the final desired result. MOSAIC is designed as a training-free framework with specially designed agents to self-reflect, create the rationale, code, and debug within a student-teacher paradigm to address the challenges of scientific code generation. This design facilitates stepwise problem decomposition, targeted error correction, and, when combined with our Consolidated Context Window (CCW), mitigates LLM hallucinations when solving complex scientific tasks involving chained subproblems. We evaluate MOSAIC on scientific coding benchmarks and demonstrate that our specialized agentic framework outperforms existing approaches in terms of accuracy, robustness, and interpretability.
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。