DynaThink: Fast or Slow? A Dynamic Decision-Making Framework for Large Language Models

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Pan, Jiabao, Zhang, Yan, Zhang, Chen, Liu, Zuozhu, Wang, Hongwei, Li, Haizhou
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866911938825420800
author Pan, Jiabao
Zhang, Yan
Zhang, Chen
Liu, Zuozhu
Wang, Hongwei
Li, Haizhou
author_facet Pan, Jiabao
Zhang, Yan
Zhang, Chen
Liu, Zuozhu
Wang, Hongwei
Li, Haizhou
contents Large language models (LLMs) have demonstrated emergent capabilities across diverse reasoning tasks via popular Chains-of-Thought (COT) prompting. However, such a simple and fast COT approach often encounters limitations in dealing with complicated problems, while a thorough method, which considers multiple reasoning pathways and verifies each step carefully, results in slower inference. This paper addresses the challenge of enabling LLMs to autonomously select between fast and slow inference methods, thereby optimizing both efficiency and effectiveness. We introduce a dynamic decision-making framework that categorizes tasks into two distinct pathways: 'Fast', designated for tasks where the LLM quickly identifies a high-confidence solution, and 'Slow', allocated for tasks that the LLM perceives as complex and for which it has low confidence in immediate solutions as well as requiring more reasoning paths to verify. Experiments on five popular reasoning benchmarks demonstrated the superiority of the DynaThink over baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2407_01009
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DynaThink: Fast or Slow? A Dynamic Decision-Making Framework for Large Language Models
Pan, Jiabao
Zhang, Yan
Zhang, Chen
Liu, Zuozhu
Wang, Hongwei
Li, Haizhou
Computation and Language
Large language models (LLMs) have demonstrated emergent capabilities across diverse reasoning tasks via popular Chains-of-Thought (COT) prompting. However, such a simple and fast COT approach often encounters limitations in dealing with complicated problems, while a thorough method, which considers multiple reasoning pathways and verifies each step carefully, results in slower inference. This paper addresses the challenge of enabling LLMs to autonomously select between fast and slow inference methods, thereby optimizing both efficiency and effectiveness. We introduce a dynamic decision-making framework that categorizes tasks into two distinct pathways: 'Fast', designated for tasks where the LLM quickly identifies a high-confidence solution, and 'Slow', allocated for tasks that the LLM perceives as complex and for which it has low confidence in immediate solutions as well as requiring more reasoning paths to verify. Experiments on five popular reasoning benchmarks demonstrated the superiority of the DynaThink over baselines.
title DynaThink: Fast or Slow? A Dynamic Decision-Making Framework for Large Language Models
topic Computation and Language
url https://arxiv.org/abs/2407.01009