Sirvi Autor "Udras, Karel" järgi

Nüüd näidatakse 1 - 1 1

Automated Resolution of Issue Reports using LLM Agents: A Systematic Literature Review
(Tartu Ülikool, 2025) Udras, Karel; Shah, Faiz Ali, juhendaja; Tartu Ülikool. Loodus- ja täppisteaduste valdkond; Tartu Ülikool. Arvutiteaduse instituut
The goal of this thesis is to evaluate the ability of large language model agents in resolving issue reports present in repositories. A systematic literature review following Kitchenham’s methodology is conducted. Systematic literature review covers studies published in 2023 and later. Articles for this systematic literature review were selected according to their relevance towards large language model agents designed for issue resolution. The systematic literature review identifies common agent designs, evaluation datasets and benchmarks, metrics, and factors influencing their performance. Findings from this thesis highlight the strengths and limitations of current large language model agents and their benchmarks.