Towards better Human-Agent Alignment: Assessing Task Utility in LLM-Powered Applications

TL;DR

This paper introduces AgentEval, a framework for assessing the utility of LLM-powered applications by automatically proposing criteria to verify alignment with user needs, aiming to improve human-agent collaboration.

Contribution

The paper presents AgentEval, a novel framework that simplifies utility verification of LLM applications through automatic criteria generation and comprehensive utility assessment.

Findings

AgentEval effectively proposes tailored utility criteria.

The framework enables robust assessment of application utility.

Preliminary analysis shows promising results in utility verification.

Abstract

The rapid development in the field of Large Language Models (LLMs) has led to a surge in applications that facilitate collaboration among multiple agents to assist humans in their daily tasks. However, a significant gap remains in assessing whether LLM-powered applications genuinely enhance user experience and task execution efficiency. This highlights the pressing need for methods to verify utility of LLM-powered applications, particularly by ensuring alignment between the application's functionality and end-user needs. We introduce AgentEval provides an implementation for the math problems, a novel framework designed to simplify the utility verification process by automatically proposing a set of criteria tailored to the unique purpose of any given application. This allows for a comprehensive assessment, quantifying the utility of an application against the suggested criteria. We present a comprehensive analysis of the robustness of quantifier's work.