DeLLMa: Decision Making Under Uncertainty with Large Language Models

TL;DR

DeLLMa is a framework that improves decision-making accuracy of large language models under uncertainty by using multi-step reasoning and principles from decision theory, achieving up to 40% better accuracy in complex tasks.

Contribution

We introduce DeLLMa, a novel multi-step reasoning framework that enhances LLM decision-making under uncertainty, integrating decision and utility theories for improved accuracy.

Findings

DeLLMa achieves up to 40% accuracy improvement over existing methods.

Performance improves with increased compute during testing.

Human evaluations validate component effectiveness.

Abstract

The potential of large language models (LLMs) as decision support tools is increasingly being explored in fields such as business, engineering, and medicine, which often face challenging tasks of decision-making under uncertainty. In this paper, we show that directly prompting LLMs on these types of decision-making problems can yield poor results, especially as the problem complexity increases. To aid in these tasks, we propose DeLLMa (Decision-making Large Language Model assistant), a framework designed to enhance decision-making accuracy in uncertain environments. DeLLMa involves a multi-step reasoning procedure that integrates recent best practices in scaling inference-time reasoning, drawing upon principles from decision theory and utility theory, to provide an accurate and human-auditable decision-making process. We validate our procedure on multiple realistic decision-making environments, demonstrating that DeLLMa can consistently enhance the decision-making performance of leading language models, and achieve up to a 40% increase in accuracy over competing methods. Additionally, we show how performance improves when scaling compute at test time, and carry out human evaluations to benchmark components of DeLLMa.