![[2602.13937] A Multi-Agent Framework for Code-Guided, Modular, and Verifiable Automated Machine Learning](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.13937] A Multi-Agent Framework for Code-Guided, Modular, and Verifiable Automated Machine Learning
Summary
The paper presents iML, a multi-agent framework for automated machine learning that enhances transparency and modularity, addressing limitations of traditional AutoML systems.
Why It Matters
This research is significant as it proposes a solution to the black-box nature of traditional AutoML frameworks, improving reliability and transparency in machine learning applications. By introducing a modular approach, it paves the way for more robust and verifiable AI systems, which is crucial for real-world engineering tasks.
Key Takeaways
- iML framework shifts AutoML from black-box to code-guided and modular.
- Introduces Code-Guided Planning to reduce hallucination in AI outputs.
- Demonstrates superior performance in real-world benchmarks compared to existing methods.
- Maintains a high success rate even with limited task descriptions.
- Enhances reliability in automated machine learning applications.
Computer Science > Machine Learning arXiv:2602.13937 (cs) [Submitted on 15 Feb 2026] Title:A Multi-Agent Framework for Code-Guided, Modular, and Verifiable Automated Machine Learning Authors:Dat Le, Duc-Cuong Le, Anh-Son Nguyen, Tuan-Dung Bui, Thu-Trang Nguyen, Son Nguyen, Hieu Dinh Vo View a PDF of the paper titled A Multi-Agent Framework for Code-Guided, Modular, and Verifiable Automated Machine Learning, by Dat Le and 6 other authors View PDF HTML (experimental) Abstract:Automated Machine Learning (AutoML) has revolutionized the development of data-driven solutions; however, traditional frameworks often function as "black boxes", lacking the flexibility and transparency required for complex, real-world engineering tasks. Recent Large Language Model (LLM)-based agents have shifted toward code-driven approaches. However, they frequently suffer from hallucinated logic and logic entanglement, where monolithic code generation leads to unrecoverable runtime failures. In this paper, we present iML, a novel multi-agent framework designed to shift AutoML from black-box prompting to a code-guided, modular, and verifiable architectural paradigm. iML introduces three main ideas: (1) Code-Guided Planning, which synthesizes a strategic blueprint grounded in autonomous empirical profiling to eliminate hallucination; (2) Code-Modular Implementation, which decouples preprocessing and modeling into specialized components governed by strict interface contracts; and (3) Code-Verifiable Int...
