相关论文: Activation Functions Considered Harmful: Recoverin…

Backdoor attacks on deep learning represent a recent threat that has gained significant attention in the research community. Backdoor defenses are mainly based on backdoor inversion, which has been shown to be generic, model-agnostic, and…

Modern CPU architectures offer strong isolation guarantees towards user applications in the form of enclaves. For instance, Intel's threat model for SGX assumes fully trusted enclaves, yet there is an ongoing debate on whether this threat…

The machine learning problem of extracting neural network parameters has been proposed for nearly three decades. Functionally equivalent extraction is a crucial goal for research on this problem. When the adversary has access to the raw…

We describe a novel family of models of multi- layer feedforward neural networks in which the activation functions are encoded via penalties in the training problem. Our approach is based on representing a non-decreasing activation function…

We propose and analyze a new family of algorithms for training neural networks with ReLU activations. Our algorithms are based on the technique of alternating minimization: estimating the activation patterns of each ReLU for all given…

Wide deployment of machine learning models on edge devices has rendered the model intellectual property (IP) and data privacy vulnerable. We propose GNNVault, the first secure Graph Neural Network (GNN) deployment strategy based on Trusted…

Intel Software Guard Extensions (SGX) provides a trusted execution environment (TEE) to run code and operate sensitive data. SGX provides runtime hardware protection where both code and data are protected even if other code components are…

Binarized Neural Networks (BNNs) are a class of deep neural networks designed to utilize minimal computational resources, which drives their popularity across various applications. Recent studies highlight the potential of mapping BNN model…

Neural network pruning is essential for reducing model complexity to enable deployment on resource constrained hardware. While performance loss of pruned networks is often attributed to the removal of critical parameters, we identify signal…

Deploying deep neural networks (DNNs) in real-world environments poses challenges due to faults that can manifest in physical hardware from radiation, aging, and temperature fluctuations. To address this, previous works have focused on…

As power consumption from AI training and inference continues to increase, AI accelerators are being integrated directly into the CPU. Intel's Advanced Matrix Extensions (AMX) is one such example, debuting on the 4th generation Intel Xeon…

Intel Software Guard Extensions (SGX) enables user-level code to create private memory regions called enclaves, whose code and data are protected by the CPU from software and hardware attacks outside the enclaves. Recent work introduces…

We here study whether training biases can make hidden neurons specialize in minimal one-hidden-layer MLPs, and whether such specialization improves prototype-based reconstruction of the training dataset from the learned weights. We consider…

Binarized Neural Networks (BNNs) deployed on memristive crossbar arrays provide energy-efficient solutions for edge computing but are susceptible to physical attacks due to memristor nonvolatility. Recently, Rajendran et al. (IEEE Embedded…

Machine learning models have been shown to leak information violating the privacy of their training set. We focus on membership inference attacks on machine learning models which aim to determine whether a data point was used to train the…

Deep neural networks are vulnerable to backdoor attacks (Trojans), where an attacker poisons the training set with backdoor triggers so that the neural network learns to classify test-time triggers to the attacker's designated target class.…

Gradient-based neural network training traditionally enforces symmetry between forward and backward propagation, requiring activation functions to be differentiable (or sub-differentiable) and strictly monotonic in certain regions to…

Modern processors can offer hardware primitives that allow a process to run in isolation. These primitives implement a trusted execution environment (TEE) in which a program can run such that the integrity and confidentiality of its…

The deployment of large language models (LLMs) on third-party devices requires new ways to protect model intellectual property. While Trusted Execution Environments (TEEs) offer a promising solution, their performance limits can lead to a…

The scope of research in the domain of activation functions remains limited and centered around improving the ease of optimization or generalization quality of neural networks (NNs). However, to develop a deeper understanding of deep…