Certifiers make neural networks vulnerable to availability attacks
To achieve reliable, robust, and safe AI systems, it is vital to implement fallback strategies when AI predictions cannot be trusted. Certifiers for neural networks are a reliable way to check the robustness of these predictions. They guarantee for some predictions that a certain class of manipulati...
Príomhchruthaitheoirí: | Lorenz, T, Kwiatkowska, M, Fritz, M |
---|---|
Formáid: | Conference item |
Teanga: | English |
Foilsithe / Cruthaithe: |
Association for Computing Machinery
2023
|
Míreanna comhchosúla
-
FullCert: deterministic end-to-end certification for training and inference of neural networks
de réir: Lorenz, T, et al.
Foilsithe / Cruthaithe: (2024) -
Certified Robustness to Text Adversarial Attacks by Randomized [MASK]
de réir: Jiehang Zeng, et al.
Foilsithe / Cruthaithe: (2023-06-01) -
Attack Vulnerability of Network Controllability.
de réir: Zhe-Ming Lu, et al.
Foilsithe / Cruthaithe: (2016-01-01) -
Bayesian inference with certifiable adversarial robustness
de réir: Wicker, M, et al.
Foilsithe / Cruthaithe: (2021) -
Vulnerability analysis on noise-injection based hardware attack on deep neural networks
de réir: Liu, Wenye, et al.
Foilsithe / Cruthaithe: (2020)