MOSAIC：多智能体协作用于任务智能的科学编码
MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific Coding
AI总结：
MOSAIC通过多智能体协作解决复杂科学编码任务，采用自反思和学生-教师模式提升准确性和鲁棒性。
摘要：
我们提出了MOSAIC，一种多智能体大型语言模型框架，用于解决具有挑战性的科学编码任务。与通用编程不同，科学工作流需要严谨、与深度领域知识相连的算法，以及领域特定推理和无需I/O测试用例的算法迭代。许多科学问题需要解决一系列子问题，最终得到期望结果。MOSAIC设计为无训练框架，包含专门设计的智能体，通过学生-教师模式自我反思、生成推理、代码和调试，以应对科学代码生成的挑战。该设计促进了逐步问题分解、目标性错误纠正，并在结合我们的整合上下文窗口（CCW）时，减轻了在解决涉及链式子问题的复杂科学任务时LLM幻觉的问题。我们在科学编码基准上评估了MOSAIC，并证明我们的专用智能体框架在准确性、鲁棒性和可解释性方面优于现有方法。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
