Smart Contract Vulnerability Auditor Using ML Models: A Literature Survey

The rising use of smart contracts in various industries makes their protection more and more important. Conventional vulnerability detection techniques have difficulty scaling and adapting since they frequently rely on rigid criteria. This paper explores how cutting-edge technology like blockchain integration, machine learning, and graph neural networks are revolutionizing the field of smart contract vulnerability detection. It also dives into fresh and interesting advances in this area. Attention-based machine learning models, GNNs, and multi-task learning models are becoming game changers with substantial potential to improve the security of smart contracts. This study aims to summarize these novel approaches and emphasize their importance to enhancing the ecosystem's overall security for smart contracts. In the end, it aims to provide a thorough framework that makes use of these strategies’ advantages to create a more stable and secure future for smart.