Search papers, labs, and topics across Lattice.
This paper addresses the limitations of existing binary code representation learning methods that primarily focus on function-level embeddings, which overlook fine-grained instruction-level relationships. By leveraging instruction alignment knowledge, the authors propose a training approach that incorporates this alignment as an auxiliary objective, leading to improved accuracy in binary code retrieval tasks. The results indicate that models fine-tuned with instruction alignment not only enhance function-level embedding quality but also yield more interpretable and discriminative representations.
Instruction alignment can significantly boost the accuracy of binary code representation learning, revealing deeper semantic connections than traditional methods.
Binary code representation learning is a fundamental problem in software security and reverse engineering. Existing methods mainly learn function-level embeddings that capture coarse-grained semantic relationships between binary functions, but they largely ignore fine-grained instruction-level correspondences. This limitation misses valuable supervision signals available from compiler debug information, which can support the learning of more accurate and interpretable binary code representations. We propose to leverage instruction alignment knowledge to further improve binary code representation learning. Our preliminary study reveals that models finetuned for function-level binary code similarity exhibit substantially better instruction alignment than their pre-trained model, suggesting a strong correlation between instruction alignment and function-level embedding quality. Motivated by this observation, we design a training approach that explicitly incorporates instruction alignment as an auxiliary training objective. Our experiments show that instruction alignment training improves retrieval accuracy and provides more discriminative signal for the model's similarity judgments.
