Assessing Code Understanding in LLMs

Cosimo Laneve, Alvise Spanò, Dalila Ressi +2

We present an empirical evaluation of Large Language Models in code understanding associated with non-trivial, semantic-preserving program transformations such as copy propagation…