<ul><li>Federated Learning (FL) allows model training across edge devices in a privacy-friendly manner.</li><li>Tool-aided Evolutionary Large Language Model (T-ELLM) framework is proposed for efficient device selection in a wireless FL environment.</li><li>T-ELLM uses natural language-based prompts to improve generalization across network conditions.</li><li>Experimental results show T-ELLM performs better in energy efficiency and adapts well to environmental changes.</li></ul>

Tool-Aided Evolutionary LLM for Generative Policy Toward Efficient Resource Management in Wireless Federated Learning

Discover more