An Empirical Study on Capability of Large Language Models in Understanding Code Semantics | Read Paper on Bytez