A Performance Study of LLM-Generated Code on Leetcode

Tristan Coignion (1) , Clément Quinton (1) , Romain Rouvoy (1)

1 SPIRALS - Self-adaptation for distributed services and large software systems

This study evaluates the efficiency of code generation by Large Language Models (LLMs) and measures their performance against human-crafted solutions using a dataset from Leetcode. We compare 18 LLMs, considering factors such as model temperature and success rate, and their impact on code performance. This research introduces a novel method for measuring and comparing the speed of LLM-generated code, revealing that LLMs produce code with comparable performance, irrespective of the adopted LLM. We also find that LLMs are capable of generating code that is, on average, more efficient than the code written by humans. The paper further discusses the use of Leetcode as a benchmarking dataset, the limita- tions imposed by potential data contamination, and the platform’s measurement reliability. We believe that our findings contribute to a better understanding of LLM capabilities in code generation and set the stage for future optimizations in the field.

Mots clés: LLM; Large Language Model; Leetcode; Performance; Python

Domaines: Génie logiciel [cs.SE]; Informatique [cs]

Tristan Coignion, Clément Quinton, Romain Rouvoy. A Performance Study of LLM-Generated Code on Leetcode. EASE'24 - 28th International Conference on Evaluation and Assessment in Software Engineering, Jun 2024, Salerno, Italy. ⟨hal-04525620⟩