Blockwise p-Tampering Attacks on Cryptographic Primitives, Extractors, and Learners
By Saeed Mahloujifar, Mohammad Mahmoody. 2017
29 citations recorded by the source corpus.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Dimitris Tsipras
- Micah Goldblum
- Bo Li
- Nicholas Carlini
- Alexander Turner
- Xinyun Chen
- Tom Goldstein
- Dawn Song
Citations in the corpus
- The Curse of Concentration in Robust Learning: Evasion and Poisoning Attacks from Concentration of Measure (2018, methodology): “However, there is a big distinction between our attacks here and those of (Mahloujifar, Diochnos, and Mahmoody 2018), as our attackers need to know the entire training sequence before tampering with them, while the attacks of Mahloujifar et al. were online.”
- Multi-party Poisoning through Generalized p-Tampering (2018, methodology): “p-tampering model for which this equivalence is not true. In fact, optimal bounds achievable by bitwise p-tampering as proved in [ACM+17] are impossible to achieve in the blockwise ptampering setting [MM17]. Despite this simplification, we still prefer to use a blockwise presentation of the random pr”
- Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses (2020, background): “Mahloujifar and Mahmoody [66] consider the setting of targeted poisoning and expand the idea to block-wise p-tampering, where entire samples in the training set of a learning pipeline can”
- Learning under p-Tampering Attacks (2018, background): “After obtaining biasing attacks, (Mahloujifar and Mahmoody, 2017) derived their ptampering targeted poisoning attacks from them by biasing the average of the loss function Loss(h(x), y) where h is the learned hypothesis and (x, y) = d is the target test.”
- Label-Consistent Backdoor Attacks (2019, background): “Attacks restricted to only using correctly label poisoned samples have been explored in prior work, being referred to as “defensible” (Mahloujifar & Mahmoody, 2017; Mahloujifar et al., 2017), “plausible” (Mahloujifar et al., 2018; Mahloujifar & Mahmoody, 2018), “visually indistinguishable” (Koh &…”