Fleet of Agents: Coordinated Problem Solving with Large Language Models (2405.06691v3)
Abstract: While numerous frameworks have been developed to enhance the reasoning abilities of 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.
- Graph of thoughts: Solving elaborate problems with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence (2024), vol. 38, pp. 17682–17690.
- Language models are few-shot learners. In Advances in Neural Information Processing Systems (2020), H. Larochelle, M. Ranzato, R. Hadsell, M. Balcan, and H. Lin, Eds., vol. 33, Curran Associates, Inc., pp. 1877–1901.
- Chase, H. Langchain. https://github.com/hwchase17/langchain, 2022.
- Daniel, K. Thinking, fast and slow. 2017.
- Evolutionary algorithms in engineering applications. Springer Science & Business Media, 2013.
- Holland, J. H. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence. MIT press, 1992.
- Metagpt: Meta programming for multi-agent collaborative framework. CoRR abs/2308.00352 (2023).
- Flows: Building blocks of reasoning and collaborating ai. arXiv preprint arXiv:2308.01285 (2023).
- Large language models are zero-shot reasoners. In Advances in Neural Information Processing Systems (2022), S. Koyejo, S. Mohamed, A. Agarwal, D. Belgrave, K. Cho, and A. Oh, Eds., vol. 35, Curran Associates, Inc., pp. 22199–22213.
- Positioning error compensation of an industrial robot using neural networks and experimental study. Chinese Journal of Aeronautics 35, 2 (2022), 346–360. https://www.sciencedirect.com/science/article/abs/pii/S1000936121001187 ; https://www.sciencedirect.com/science/article/pii/S1000936121001187.
- Camel: Communicative agents for" mind" exploration of large scale language model society. arXiv preprint arXiv:2303.17760 (2023).
- More agents is all you need, 2024.
- Chameleon: Plug-and-play compositional reasoning with large language models. ArXiv abs/2304.09842 (2023).
- Self-refine: Iterative refinement with self-feedback. arXiv preprint arXiv:2303.17651 (2023).
- A hybrid genetic - particle swarm optimization algorithm for the vehicle routing problem. Expert Systems with Applications 37, 2 (2010), 1446–1455. https://dblp.uni-trier.de/db/journals/eswa/eswa37.html#MarinakisM10 ; https://www.sciencedirect.com/science/article/pii/S0957417409006460 ; https://dl.acm.org/doi/10.1016/j.eswa.2009.06.085.
- Explaining machine learning classifiers through diverse counterfactual explanations. In Proceedings of the 2020 conference on fairness, accountability, and transparency (2020), pp. 607–617.
- Nakajima, Y. Babyagi. https://github.com/yoheinakajima/babyagi, 2023.
- Show your work: Scratchpads for intermediate computation with language models. CoRR abs/2112.00114 (2021).
- Refiner: Reasoning feedback on intermediate representations. arXiv preprint arXiv:2304.01904 (2023).
- Richards, T. B. Autogpt. https://github.com/Significant-Gravitas/Auto-GPT, 2023.
- Dynamic optimization of fuzzy cognitive maps for time series forecasting. Knowledge-Based Systems 105 (2016), 29–37. https://dl.acm.org/doi/10.1016/j.knosys.2016.04.023 ; https://dblp.uni-trier.de/db/journals/kbs/kbs105.html#SalmeronF16 ; https://dx.doi.org/10.1016/j.knosys.2016.04.023 ; http://dx.doi.org/10.1016/j.knosys.2016.04.023 ; https://www.sciencedirect.com/science/article/pii/S0950705116300752.
- Hugginggpt: Solving ai tasks with chatgpt and its friends in huggingface. ArXiv abs/2303.17580 (2023).
- Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171 (2022).
- Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems 35 (2022), 24824–24837.
- The multidimensional wisdom of crowds. Advances in neural information processing systems 23 (2010).
- Autogen: Enabling next-gen LLM applications via multi-agent conversation framework. CoRR abs/2308.08155 (2023).
- Tree of thoughts: Deliberate problem solving with large language models. Advances in Neural Information Processing Systems 36 (2024).
- React: Synergizing reasoning and acting in language models. In The Eleventh International Conference on Learning Representations (2023).
Collections
Sign up for free to add this paper to one or more collections.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.