Salus’s research extends beyond smart contracts, proposing that the Lightning Cat technology can be applied to identify weaknesses in other types of code throughout the entire blockchain stack.
Salus, a leading Web3 security company, has accomplished a big milestone with the publication of its research in ‘Scientific Reports’, a sub-journal of the esteemed ‘Nature.’ The research paper “Deep learning-based solution for smart contract vulnerabilities detection” explores the potential of deep learning in identifying vulnerabilities in smart contracts.
Deep Learning for Smart Contract Security
Web3, often referred to as the decentralized web, is a transformative evolution of the internet that leverages blockchain technology. As the decentralized ecosystem continues to gain prominence, robust security mechanisms become paramount.
Smart contracts, self-executing contracts with the terms of the agreement written directly into code, play a central role in many Decentralized Applications (DApps) and blockchain networks. However, ensuring the security and integrity of these smart contracts has posed a considerable challenge.
The core of Salus’s research revolves around the development of a deep learning-based solution named “Lightning Cat” designed for smart contract vulnerability detection. This innovative approach employs three deep learning models to maximize the identification of vulnerabilities within smart contracts, achieving an impressive f1-score of 93.53%.
This remarkable result positions Salus’s solution as a potentially superior method for identifying vulnerabilities compared to existing approaches. Traditional vulnerability detection methods often struggle with false positives or negatives, particularly in the intricate logic of smart contract code.
Salus’s deep learning solution, Lightning Cat, breaks free from the limitations of predefined rules. It is adaptable enough to constantly learn and update itself, making it more capable of identifying new and changing attack vectors.
Salus’s research extends beyond smart contracts, proposing that the Lightning Cat technology can be applied to identify weaknesses in other types of code throughout the entire blockchain stack. This versatility enhances code dependability and reduces the risk of exploitation, potentially preventing significant financial losses.
Jiayi Li, Co-Founder of Salus, expressed gratitude for the recognition by ‘Scientific Reports’ stating:
“We’re grateful to ‘Scientific Reports,’ a sub-journal of ‘Nature,’ for publishing our research into the mitigation of smart contract vulnerabilities through deep learning. This decision is a testament not only to the researchers’ methodology but to the progression of Web3.”
Web3 Security Development
While Salus has taken the initiative for Web3 security development, other players in the space are also making strides. DeFi Protocol Frontier, for instance, has incorporated a fraud-detection engine in its in-browser wallet, safeguarding users from malicious transactions.
In the broader Web3 ecosystem, Animoca Brands, Amazon Web Services (AWS), and Polygon Labs have joined forces to support Web3 builders through the MoonRealm Express Accelerator, offering resources such as masterclasses, hackathons, and project incubation.
Additionally, the TON Foundation has partnered with Chainbase and Tencent Cloud to simplify blockchain development for Web3 mass adoption across the Asia-Pacific region. Tencent Cloud will provide computing resources and network connectivity to support TON validators and facilitate the development of web applications and bots within Telegram.
As Web3 evolves, these alliances and advancements in security signal a promising future for the decentralized and secure internet of the next generation.
next
Source: https://www.coinspeaker.com/salus-web3-security-deep-learning/