Saved in:
Bibliographic Details
Main Authors: Chari, Anirudh, Tiwari, Aditya, Lian, Richard, Reddy, Suraj, Zhou, Brian
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.19278
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917907699597312
author Chari, Anirudh
Tiwari, Aditya
Lian, Richard
Reddy, Suraj
Zhou, Brian
author_facet Chari, Anirudh
Tiwari, Aditya
Lian, Richard
Reddy, Suraj
Zhou, Brian
contents Large Language Models (LLMs) have demonstrated remarkable reasoning capabilities through chain-of-thought prompting, yet discovering effective reasoning methods for complex problems remains challenging due to the vast space of possible intermediate steps. We introduce Ant Colony Optimization-guided Tree of Thought (ACO-ToT), a novel algorithm that combines ACO with LLMs to discover optimal reasoning paths for complex problems efficiently. Drawing inspiration from Hebbian learning in neurological systems, our method employs a collection of distinctly fine-tuned LLM "ants" to traverse and lay pheromone trails through a centralized tree of thought, with each ant's movement governed by a weighted combination of existing pheromone trails and its own specialized expertise. The algorithm evaluates complete reasoning paths using a mixture-of-experts-based scoring function, with pheromones reinforcing productive reasoning paths across iterations. Experiments on three challenging reasoning tasks (GSM8K, ARC-Challenge, and MATH) demonstrate that ACO-ToT performs significantly better than existing chain-of-thought optimization approaches, suggesting that incorporating biologically inspired collective search mechanisms into LLM inference can substantially enhance reasoning capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2501_19278
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Pheromone-based Learning of Optimal Reasoning Paths
Chari, Anirudh
Tiwari, Aditya
Lian, Richard
Reddy, Suraj
Zhou, Brian
Computation and Language
Large Language Models (LLMs) have demonstrated remarkable reasoning capabilities through chain-of-thought prompting, yet discovering effective reasoning methods for complex problems remains challenging due to the vast space of possible intermediate steps. We introduce Ant Colony Optimization-guided Tree of Thought (ACO-ToT), a novel algorithm that combines ACO with LLMs to discover optimal reasoning paths for complex problems efficiently. Drawing inspiration from Hebbian learning in neurological systems, our method employs a collection of distinctly fine-tuned LLM "ants" to traverse and lay pheromone trails through a centralized tree of thought, with each ant's movement governed by a weighted combination of existing pheromone trails and its own specialized expertise. The algorithm evaluates complete reasoning paths using a mixture-of-experts-based scoring function, with pheromones reinforcing productive reasoning paths across iterations. Experiments on three challenging reasoning tasks (GSM8K, ARC-Challenge, and MATH) demonstrate that ACO-ToT performs significantly better than existing chain-of-thought optimization approaches, suggesting that incorporating biologically inspired collective search mechanisms into LLM inference can substantially enhance reasoning capabilities.
title Pheromone-based Learning of Optimal Reasoning Paths
topic Computation and Language
url https://arxiv.org/abs/2501.19278