Please use this identifier to cite or link to this item: https://rd.uffs.edu.br/handle/prefix/9687
Full metadata record
DC FieldValueLanguage
dc.contributor.advisor1Feitosa, Samuel da Silva-
dc.creatorMachado, Heitor Fernandes Vicente-
dc.date2026-06-24-
dc.date.accessioned2026-10-09T09:41:29Z-
dc.date.available2026-
dc.date.available2026-10-09T09:41:29Z-
dc.date.issued2026-
dc.identifier.urihttps://rd.uffs.edu.br/handle/prefix/9687-
dc.description.resumoLarge Language Models have shown promising results in automatic code generation, but challenges still exist related to the quality of the produced implementations and the effective use of testing during the generation process. This work investigates the use of unit tests automatically generated by LLMs in contrast with manually crafted tests as an iterative code refinement mechanism for programming problems. The proposed approach consists of a tool capable of generating unit tests from a task specification, validating imple- mentations produced by the models, and performing successive correction cycles guided by the test results. The methodology was evaluated using the HumanEval benchmark in conjunction with the EvalPlus framework, considering scenarios of direct generation, self-validation with automatically generated tests, and validation using the benchmark’s original tests. The evaluation considered the Pass@1 and Pass@10 metrics to measure problem-solving perfor- mance, as well as the analysis of the generated test suites, classified as valid, invalid, or not produced. The results demonstrate that synthetic tests are capable of improving the performance of all evaluated models, although they still present lower performance compared to that obtained with manually crafted tests. The findings indicate that automatic test generation by LLMs represents a viable strategy to support code generation and refinement processes.pt_BR
dc.description.provenanceSubmitted by Biblioteca Chapeco (biblio.ch@uffs.edu.br) on 2026-10-06T12:06:38Z No. of bitstreams: 1 MACHADO.pdf: 659768 bytes, checksum: 7c91f7247f8309724d19fd224a8223a2 (MD5)en
dc.description.provenanceApproved for entry into archive by Franciele Scaglioni da Cruz (franciele.cruz@uffs.edu.br) on 2026-10-09T09:41:29Z (GMT) No. of bitstreams: 1 MACHADO.pdf: 659768 bytes, checksum: 7c91f7247f8309724d19fd224a8223a2 (MD5)en
dc.description.provenanceMade available in DSpace on 2026-10-09T09:41:29Z (GMT). No. of bitstreams: 1 MACHADO.pdf: 659768 bytes, checksum: 7c91f7247f8309724d19fd224a8223a2 (MD5) Previous issue date: 2026en
dc.languageporpt_BR
dc.publisherUniversidade Federal da Fronteira Sulpt_BR
dc.publisher.countryBrasilpt_BR
dc.publisher.departmentCampus Chapecópt_BR
dc.publisher.initialsUFFSpt_BR
dc.rightsAcesso Abertopt_BR
dc.subjectInteligência artificialpt_BR
dc.subjectGeração de códigopt_BR
dc.subjectTestespt_BR
dc.subjectEngenharia de softwarept_BR
dc.titleEvaluating LLM-generated and human-written unit tests for iterative code refinementpt_BR
dc.typeMonografiapt_BR
Appears in Collections:Ciência da Computação

Files in This Item:
File Description SizeFormat 
MACHADO.pdf644.3 kBAdobe PDFView/Open
Show simple item record Recommend this item


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.