MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific Coding
This addresses the challenge of rigorous and interconnected scientific coding workflows for researchers, though it appears incremental as it builds on existing multi-agent and LLM methods.
The authors tackled the problem of generating scientific code by introducing MOSAIC, a multi-agent LLM framework that uses a student-teacher paradigm and a Consolidated Context Window to improve accuracy, robustness, and interpretability on scientific coding benchmarks.
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.