Work

A Causal Perspective on Measuring, Explaining and Mitigating Smells in LLM-Generated Code

Year: 2026

Type: conference-paper

Abstract: Recent advances in large language models (LLMs) have accelerated their adoption in software engineering contexts. However, concerns persist about the structural quality of the code they produce. In particular, LLMs often replicate poor coding practices, introducing code smells (i.e., patterns that hinder readability, maintainability, or design integrity). Although prior research has examined the detection or repair of smells, we still lack a clear understanding of how and when these issues emerge in generated code.

Source: IEEE/ACM International Conference on Software Engineering (ICSE)

DOI: 10.1145/3744916.3773164

Authors: Alejandro Velasco, Daniel Rodriguez-Cardenas, Dipin Khati, David N. Palacio, Lutfar Rahman Alif, Denys Poshyvanyk

Institutions: William & Mary, Microsoft (United States), University of Dhaka

Topic: Software Engineering Research

Subfield: Information Systems

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funders: National Science Foundation, Cisco Systems

Awards: Collaborative Research: SHF: Medium: Toward Understandability and Interpretability for Neural Language Models of Source Code, Planning: CIRC: Planning-C: A Holistic Benchmark for Accelerating Research in Large Language Models for Software Engineering

OpenAlex ID: W7106254141

FWCI:

Cites:

Cited by:

Related to: