English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Deep neural networks (DNNs) are widely deployed on real-world devices.Concerns regarding their security have gained great attention from researchers.Recently, a new weight modification attack called bit flip attack (BFA) wasproposed, which exploits memory fault inject techniques such as row hammer toattack quantized models in the deployment stage. With only a few bit flips, thetarget model can be rendered useless as a random guesser or even be implantedwith malicious functionalities. In this work, we seek to further reduce thenumber of bit flips. We propose a training-assisted bit flip attack, in whichthe adversary is involved in the training stage to build a high-risk model torelease. This high-risk model, obtained coupled with a corresponding maliciousmodel, behaves normally and can escape various detection methods. The resultson benchmark datasets show that an adversary can easily convert this high-riskbut normal model to a malicious one on victim's side by \textbf{flipping onlyone critical bit} on average in the deployment stage. Moreover, our attackstill poses a significant threat even when defenses are employed. The codes forreproducing main experiments are available at\url{https://github.com/jianshuod/TBA}.

One-bit Flip is All You Need: When Bit-flip Attack Meets Model Training