Adversaries can create backdoored neural networks during outsourced training that maintain high accuracy on normal data but misbehave on attacker-chosen triggers.
Adversarial classification
2 Pith papers cite this work, alongside 548 external citations. Polarity classification is still indexing.
2
Pith papers citing it
548
external citations · OpenAlex
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Hamm-grams are a new class of fixed-length regular expressions over bytes with single-character wildcards, mined efficiently with LSH and clustering to yield more robust features than n-grams for malware classification and detection.
citing papers explorer
-
BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain
Adversaries can create backdoored neural networks during outsourced training that maintain high accuracy on normal data but misbehave on attacker-chosen triggers.
-
Hamm-Grams: An Algorithm for Mining Regular Expressions of Bytes
Hamm-grams are a new class of fixed-length regular expressions over bytes with single-character wildcards, mined efficiently with LSH and clustering to yield more robust features than n-grams for malware classification and detection.