Toward Architecture-Aware Evaluation Metrics for LLM Agents

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Souza, Débora, Machado, Patrícia
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866911402558488576
author Souza, Débora
Machado, Patrícia
author_facet Souza, Débora
Machado, Patrícia
contents LLM-based agents are becoming central to software engineering tasks, yet evaluating them remains fragmented and largely model-centric. Existing studies overlook how architectural components, such as planners, memory, and tool routers, shape agent behavior, limiting diagnostic power. We propose a lightweight, architecture-informed approach that links agent components to their observable behaviors and to the metrics capable of evaluating them. Our method clarifies what to measure and why, and we illustrate its application through real world agents, enabling more targeted, transparent, and actionable evaluation of LLM-based agents.
format Preprint
id arxiv_https___arxiv_org_abs_2601_19583
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Toward Architecture-Aware Evaluation Metrics for LLM Agents
Souza, Débora
Machado, Patrícia
Software Engineering
D.2.7; I.2.2; I.2.7
LLM-based agents are becoming central to software engineering tasks, yet evaluating them remains fragmented and largely model-centric. Existing studies overlook how architectural components, such as planners, memory, and tool routers, shape agent behavior, limiting diagnostic power. We propose a lightweight, architecture-informed approach that links agent components to their observable behaviors and to the metrics capable of evaluating them. Our method clarifies what to measure and why, and we illustrate its application through real world agents, enabling more targeted, transparent, and actionable evaluation of LLM-based agents.
title Toward Architecture-Aware Evaluation Metrics for LLM Agents
topic Software Engineering
D.2.7; I.2.2; I.2.7
url https://arxiv.org/abs/2601.19583