adversarial training
PulseAugur coverage of adversarial training — every cluster mentioning adversarial training across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New method enhances detection of malicious AI instructions
Researchers have developed a new method for detecting malicious instructions embedded within text, a vulnerability known as indirect prompt injection (IPI). This approach is context- and query-aware, improving upon exis…
-
ROMS-IMLE: Minimalist generative model challenges multi-step necessity
Researchers have introduced ROMS-IMLE, a novel generative model that challenges the prevailing belief in the necessity of gradual, multi-step transformations for high-quality sample generation. By adopting a minimalist …
-
New hybrid defense enhances NIDS against adversarial attacks
Researchers have developed a hybrid defense mechanism to protect Network Intrusion Detection Systems (NIDS) from adversarial attacks. This approach combines Adversarial Training (AT) and Gaussian Data Augmentation (GDA)…
-
Bilevel optimization framework detailed for Neural Architecture Search
This paper provides a structured overview of Neural Architecture Search (NAS) by framing it as a bilevel optimization problem. It categorizes existing NAS methods into sampling-based and bilevel theory-based approaches.…
-
Gradient leakage attacks threaten GNNs in circuit design
A new research paper details the first comprehensive evaluation of gradient leakage attacks (GLAs) on graph neural networks (GNNs) used in circuit design and hardware security. The study reveals that GLAs can expose sen…
-
New GRAPE framework boosts neural network adversarial robustness
Researchers have introduced GRAPE, a novel training framework designed to enhance the adversarial robustness of neural networks while maintaining compact model sizes. GRAPE distinguishes itself by treating robust model …
-
SORA method prevents catastrophic overfitting in adversarial training
Researchers have introduced SORA, a novel method for adversarial training (AT) designed to combat catastrophic overfitting in fast AT variants. SORA addresses this by formalizing Epsilon Overfitting (EO) and proposing P…
-
New theory bounds transient amplification in coupled gradient descent
Researchers have developed a new pseudospectral theory to analyze transient amplification in coupled gradient descent, a method used in bilevel optimization and adversarial training. The theory provides sharp bounds for…
-
Matching Principle unifies ML robustness with geometric theory
A new paper introduces the "Matching Principle," a geometric theory that unifies various robustness techniques in representation learning. The principle suggests that instead of treating issues like domain adaptation an…
-
New framework RobustLT tackles adversarial training on imbalanced datasets
Researchers have developed a new framework called RobustLT to improve adversarial training for deep neural networks, particularly on datasets with long-tail distributions. The framework addresses limitations in current …