Fleet of Agents: Coordinated Problem Solving with Large Language Models

TL;DR

The paper introduces Fleet of Agents (FoA), a framework that uses multiple autonomous LLM agents with a genetic particle filtering approach to improve problem-solving efficiency, balancing cost and quality effectively.

Contribution

FoA is a novel framework that enables dynamic exploration in large language models through multiple agents and heuristic resampling, achieving better cost-quality trade-offs.

Findings

FoA improves solution quality by ~5% on benchmark tasks.

FoA reduces computational cost to ~40% of state-of-the-art methods.

FoA with LLaMA3.2-11B outperforms LLaMA3.2-90B.

Abstract

While numerous frameworks have been developed to enhance the reasoning abilities of large language models (LLMs), there is a scarcity of methods that effectively balance the trade-off between cost and quality. In this paper, we introduce Fleet of Agents (FoA), a novel and intuitive yet principled framework utilizing LLMs as agents to navigate through dynamic tree searches, employing a genetic-type particle filtering approach. FoA spawns a multitude of agents, each exploring the search space autonomously, followed by a selection phase where resampling based on a heuristic value function optimizes the balance between exploration and exploitation. This mechanism enables dynamic branching, adapting the exploration strategy based on discovered solutions. We conduct extensive experiments on three benchmark tasks, ``Game of 24'', ``Mini-Crosswords'', and ``WebShop'', utilizing four different LLMs, ``GPT-3.5'', ``GPT-4'', ``LLaMA3.2-11B'', and ``LLaMA3.2-90B''. On average across all tasks and LLMs, FoA obtains a quality improvement of ~5% while requiring only ~40% of the cost of previous SOTA methods. Notably, our analyses reveal that (1) FoA achieves the best cost-quality trade-off among all benchmarked methods and (2) FoA + LLaMA3.2-11B surpasses the Llama3.2-90B model. FoA is publicly available at https://github.com/au-clan/FoA.