LLM-Based Embeddings for Program Analysis and Optimization

Recent advances have highlighted the potential of machine learning, particularly Large Language Models (LLMs), for analyzing and optimizing programs. We present the first application of program embeddings from LLMCompiler---an LLM massively pretrained on intermediate representation (IR) code---to representative program analysis and optimization tasks. We generate program embeddings directly from source and IR code using a simple approach: split programs into chunks, independently embed each chunk with pretrained LLMs, and then aggregate the chunk embeddings into a single program embedding. Our experiments show that combining source and IR code embeddings achieves an error rate of 1.54% in algorithm classification, a 12% improvement over the current state-of-the-art, and a competitive accuracy on heterogeneous device mapping. These findings suggest that training a performance-aware LLM for embedding IR code might yield state-of-the-art results in code optimization tasks.