A Multi-Agent LLM System for Automated Tuning of Specialized Neural Network Models.

arXiv:2610.13663v1 Announce Type: new Abstract: Many machine learning approaches are two-level: an expert selects a class of methods, within this class the choice of key components is critical and usually manual. This paper addresses the automatic tuning of th