MOTIF: Multi-strategy Optimization via Turn-based Interactive Framework

TL;DR

MOTIF introduces a turn-based, multi-agent framework leveraging LLMs to optimize multiple interdependent components in combinatorial optimization problems, leading to more diverse and effective solver designs.

Contribution

The paper presents MOTIF, a novel turn-based interactive framework using LLMs for joint multi-strategy optimization in solver design, expanding beyond single-element heuristics.

Findings

MOTIF outperforms existing methods across multiple COP domains.

Turn-based multi-agent prompting enhances solver component diversity.

Framework promotes emergent cooperation among optimization strategies.

Abstract

Designing effective algorithmic components remains a fundamental obstacle in tackling NP-hard combinatorial optimization problems (COPs), where solvers often rely on carefully hand-crafted strategies. Despite recent advances in using large language models (LLMs) to synthesize high-quality components, most approaches restrict the search to a single element - commonly a heuristic scoring function - thus missing broader opportunities for innovation. In this paper, we introduce a broader formulation of solver design as a multi-strategy optimization problem, which seeks to jointly improve a set of interdependent components under a unified objective. To address this, we propose Multi-strategy Optimization via Turn-based Interactive Framework (MOTIF) - a novel framework based on Monte Carlo Tree Search that facilitates turn-based optimization between two LLM agents. At each turn, an agent improves one component by leveraging the history of both its own and its opponent's prior updates, promoting both competitive pressure and emergent cooperation. This structured interaction broadens the search landscape and encourages the discovery of diverse, high-performing solutions. Experiments across multiple COP domains show that MOTIF consistently outperforms state-of-the-art methods, highlighting the promise of turn-based, multi-agent prompting for fully automated solver design.