ML Study of MaliciousTransactions in Ethereum

TL;DR

This paper introduces machine learning methods to detect malicious transactions in Ethereum by analyzing smart contract code and transaction features, achieving effective identification of malicious activities.

Contribution

It proposes two novel ML approaches using GPT2 and CodeLlama for contract analysis, and combines gas and signature features with XGBOOST for transaction detection.

Findings

GPT2-based opcode analysis effectively detects malicious contracts.

CodeLlama fine-tuned on Solidity source improves contract vulnerability detection.

XGBOOST model using gas and signature features accurately identifies malicious transactions.

Abstract

Smart contracts are a major tool in Ethereum transactions. Therefore hackers can exploit them by adding code vulnerabilities to their sources and using these vulnerabilities for performing malicious transactions. This paper presents two successful approaches for detecting malicious contracts: one uses opcode and relies on GPT2 and the other uses the Solidity source and a LORA fine-tuned CodeLlama. Finally, we present an XGBOOST model that combines gas properties and Hexa-decimal signatures for detecting malicious transactions. This approach relies on early assumptions that maliciousness is manifested by the uncommon usage of the contracts' functions and the effort to pursue the transaction.