Exploring Autonomous Agents: A Closer Look at Why They Fail When Completing Tasks

TL;DR

This paper introduces a benchmark and failure analysis framework for autonomous agents powered by LLMs, revealing key failure causes and proposing improvements to enhance robustness and task success rates.

Contribution

It presents a comprehensive benchmark, a failure taxonomy, and actionable insights to improve autonomous agent performance beyond success rate metrics.

Findings

Approximately 50% task completion rate observed.

Identified planning errors, execution issues, and response mistakes as main failure causes.

Proposed targeted improvements for planning and self-diagnosis.

Abstract

Autonomous agent systems powered by Large Language Models (LLMs) have demonstrated promising capabilities in automating complex tasks. However, current evaluations largely rely on success rates without systematically analyzing the interactions, communication mechanisms, and failure causes within these systems. To bridge this gap, we present a benchmark of 34 representative programmable tasks designed to rigorously assess autonomous agents. Using this benchmark, we evaluate three popular open-source agent frameworks combined with two LLM backbones, observing a task completion rate of approximately 50%. Through in-depth failure analysis, we develop a three-tier taxonomy of failure causes aligned with task phases, highlighting planning errors, task execution issues, and incorrect response generation. Based on these insights, we propose actionable improvements to enhance agent planning and self-diagnosis capabilities. Our failure taxonomy, together with mitigation advice, provides an empirical foundation for developing more robust and effective autonomous agent systems in the future.