EvoSpeak: Large Language Models for Interpretable Genetic Programming-Evolved Heuristics

2025-10-02

AI Summary

Genetic programming (GP) often yields heuristic rules with poor interpretability, slow convergence, and limited cross-task transferability when applied to dynamic, large-scale optimization problems. To address these limitations, this paper proposes LLM-GP, a synergistic framework integrating large language models (LLMs) with GP. Specifically, LLMs generate semantically grounded warm-start populations for GP by automatically translating natural-language problem descriptions into executable heuristic rules; further, the framework enables symbolic reasoning–based knowledge transfer and user-preference–aligned rule evolution. Evaluated on dynamic flexible job-shop scheduling, LLM-GP significantly improves both solution quality and convergence speed in single- and multi-objective settings. Moreover, it produces human-readable decision-logic reports, enhancing transparency, practical utility, and cross-task adaptability—thereby bridging the gap between automated search and domain-informed, interpretable optimization.

Technology Category

• Search and Optimization: Metareasoning and Metaheuristics
• Knowledge Representation and Reasoning: Logic Programming
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Genetic programming (GP) has demonstrated strong effectiveness in evolving tree-structured heuristics for complex optimization problems. Yet, in dynamic and large-scale scenarios, the most effective heuristics are often highly complex, hindering interpretability, slowing convergence, and limiting transferability across tasks. To address these challenges, we present EvoSpeak, a novel framework that integrates GP with large language models (LLMs) to enhance the efficiency, transparency, and adaptability of heuristic evolution. EvoSpeak learns from high-quality GP heuristics, extracts knowledge, and leverages this knowledge to (i) generate warm-start populations that accelerate convergence, (ii) translate opaque GP trees into concise natural-language explanations that foster interpretability and trust, and (iii) enable knowledge transfer and preference-aware heuristic generation across related tasks. We verify the effectiveness of EvoSpeak through extensive experiments on dynamic flexible job shop scheduling (DFJSS), under both single- and multi-objective formulations. The results demonstrate that EvoSpeak produces more effective heuristics, improves evolutionary efficiency, and delivers human-readable reports that enhance usability. By coupling the symbolic reasoning power of GP with the interpretative and generative strengths of LLMs, EvoSpeak advances the development of intelligent, transparent, and user-aligned heuristics for real-world optimization problems.

Problem

• Enhancing interpretability of complex genetic programming heuristics
• Accelerating convergence in dynamic large-scale optimization problems
• Enabling knowledge transfer across related optimization tasks

Innovation

• Integrates genetic programming with large language models
• Generates warm-start populations to accelerate convergence
• Translates GP trees into natural-language explanations