An LLM-Empowered Adaptive Evolutionary Algorithm For Multi-Component Deep Learning Systems

2025-01-01

AI Summary

Traditional multi-objective evolutionary algorithms (MOEAs) struggle to balance search efficiency and solution diversity in safety verification of multi-component deep learning (MCDL) systems. Method: This paper proposes μMOEA, the first LLM-augmented adaptive MOEA, which integrates large language models into the MOEA framework. It leverages LLMs’ semantic understanding to generate semantically rich and uniformly distributed initial populations, and employs quantitative evolutionary feedback to iteratively refine LLM inference—dynamically producing diverse, high-quality seeds to escape local optima. Contribution/Results: Evaluated on MCDL safety violation detection, μMOEA achieves a 2.3× speedup over state-of-the-art methods and improves the hypervolume (HV) metric by 37%, demonstrating both the efficacy and generalizability of LLM-driven evolutionary search for safety-critical deep learning verification.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Multi-objective evolutionary algorithms (MOEAs) are widely used for searching optimal solutions in complex multi-component applications. Traditional MOEAs for multi-component deep learning (MCDL) systems face challenges in enhancing the search efficiency while maintaining the diversity. To combat these, this paper proposes $mu$MOEA, the first LLM-empowered adaptive evolutionary search algorithm to detect safety violations in MCDL systems. Inspired by the context-understanding ability of Large Language Models (LLMs), $mu$MOEA promotes the LLM to comprehend the optimization problem and generate an initial population tailed to evolutionary objectives. Subsequently, it employs adaptive selection and variation to iteratively produce offspring, balancing the evolutionary efficiency and diversity. During the evolutionary process, to navigate away from the local optima, $mu$MOEA integrates the evolutionary experience back into the LLM. This utilization harnesses the LLM's quantitative reasoning prowess to generate differential seeds, breaking away from current optimal solutions. We evaluate $mu$MOEA in finding safety violations of MCDL systems, and compare its performance with state-of-the-art MOEA methods. Experimental results show that $mu$MOEA can significantly improve the efficiency and diversity of the evolutionary search.

Problem

• Multi-component Deep Learning
• Multi-objective Evolutionary Algorithms
• Accelerated Search with Diversity

Innovation

• Multi-objective Evolutionary Algorithm (μMOEA)
• Large Language Models (LLMs) Integration
• Multi-component Deep Learning (MCDL) Systems Optimization