Bi-Directional Transformers vs. word2vec: Discovering Vulnerabilities in Lifted Compiled Code

TL;DR

This paper compares word2vec and transformer-based embeddings like BERT and RoBERTa for vulnerability detection in compiled code, finding simpler word2vec models outperform complex transformers in limited data scenarios.

Contribution

It is the first study to compare word2vec with multiple bidirectional transformers for vulnerability detection in LLVM IR code.

Findings

Word2vec Skip-Gram achieved 92% validation accuracy.

Transformers did not outperform simpler word2vec models.

Complex contextual embeddings may not be advantageous with limited data.

Abstract

Detecting vulnerabilities within compiled binaries is challenging due to lost high-level code structures and other factors such as architectural dependencies, compilers, and optimization options. To address these obstacles, this research explores vulnerability detection using natural language processing (NLP) embedding techniques with word2vec, BERT, and RoBERTa to learn semantics from intermediate representation (LLVM IR) code. Long short-term memory (LSTM) neural networks were trained on embeddings from encoders created using approximately 48k LLVM functions from the Juliet dataset. This study is pioneering in its comparison of word2vec models with multiple bidirectional transformers (BERT, RoBERTa) embeddings built using LLVM code to train neural networks to detect vulnerabilities in compiled binaries. Word2vec Skip-Gram models achieved 92% validation accuracy in detecting vulnerabilities, outperforming word2vec Continuous Bag of Words (CBOW), BERT, and RoBERTa. This suggests that complex contextual embeddings may not provide advantages over simpler word2vec models for this task when a limited number (e.g. 48K) of data samples are used to train the bidirectional transformer-based models. The comparative results provide novel insights into selecting optimal embeddings for learning compiler-independent semantic code representations to advance machine learning detection of vulnerabilities in compiled binaries.