New flaw found in SAP defense, reducing its effectiveness to 0.1%.
problem Weakness in Stochastic Activation Pruning defense against adversarial attacks.
method Re-examined the implementation of SAP and introduced a new BPDA attack.
result SAP's effectiveness reduced to 0.1% when properly applied.
Neural networks are known to be vulnerable to adversarial examples. Carefully chosen perturbations to real images, while imperceptible to humans, induce misclassification and threaten the reliability of deep learning systems in the wild. To guard against adversarial examples, we take inspiration from game theory and ca…
A new method prunes neural networks efficiently without losing effectiveness.
problem Efficient pruning of neural networks without sacrificing performance.
method Deterministic approximation of binary gates and L0 regularization. result Pruning neural networks significantly without loss in effectiveness.
Deep neural networks have dramatically achieved great success on a variety of challenging tasks. However, most successful DNNs have an extremely complex structure, leading to extensive research on model compression.As a significant area of progress in model compression, traditional gradual pruning approaches involve an…
Pruning neural networks adds differential privacy noise, preserving data utility.
problem Achieving differential privacy in neural networks without sacrificing data utility.
method Proving equivalence between pruning and adding differential privacy noise to hidden-layer activations.
result Pruning can be a more effective alternative to adding differential privacy noise for neural networks.
Optimizes pruning masks for neural networks using probabilistic fine-tuning and PAC-Bayes bounds.
problem Improving neural network performance through adaptive pruning of weights.
method Optimizes stochastic pruning masks by minimizing expected loss, considering data-adaptive regularization and feature alignment.
result Probabilistic fine-tuning leads to improved test error over baseline methods in neural networks.
Real-time pruning during training reduces network size and training time.
problem Efficiently reducing neural network size without sacrificing accuracy.
method Activation density-based pruning during training.
result Up to 200x reduction in parameters and 60x reduction in inference compute operations.
SCBF preserves medical data privacy by training models without sharing inputs.
problem Privacy concerns in medical data collection and training.
method Stochastic Channel-Based Federated Learning (SCBF) with pruning.
result SCBF outperforms Federated Averaging with better performance and faster saturating speed.
With the rapid scaling up of deep neural networks (DNNs), extensive research studies on network model compression such as weight pruning have been performed for improving deployment efficiency. This work aims to advance the compression beyond the weights to neuron activations. We propose the joint regularization techni…
Gibbs pruning optimizes neural networks by combining physics and regularization.
problem Large neural networks are impractical for many applications.
method Combines statistical physics and stochastic regularization to train and prune networks simultaneously.
result Gibbs pruning achieves state-of-the-art performance on ResNet-56.
Sparsification is an efficient approach to accelerate CNN inference, but it is challenging to take advantage of sparsity in training procedure because the involved gradients are dynamically changed. Actually, an important observation shows that most of the activation gradients in back-propagation are very close to zero…
To improve the execution speed and efficiency of neural networks in embedded systems, it is crucial to decrease the model size and computational complexity. In addition to conventional compression techniques, e.g., weight pruning and quantization, removing unimportant activations can reduce the amount of data communica…
Sparse Transformers degrade semantic information first, with early layers encoding more.
problem Understanding how sparse Transformers affect learned representations and semantic information.
method Probed Transformers with progressively pruned weights to observe changes in semantic information and model behavior.
result Complex semantic information is first to degrade in sparse Transformers, with early layers encoding more.
A new pruning method reduces neural network computation without retraining.
problem Efficiently reduce neural network computation while maintaining accuracy.
method Structured directional pruning via perturbation orthogonal projection.
result Achieves state-of-the-art pruned accuracy without retraining.
TENP prunes experts and neurons in Mixture-of-Experts models for efficient deployment.
problem Efficient deployment of large language models constrained by static parameter footprint.
method Structured Trapezoidal ExpertNeuron Pruning (TENP) identifies and retains important experts and neurons.
result DeepSeek model achieves 10% better performance on code generation tasks with 40% expert sparsity.
New method prunes recurrent networks efficiently, improving performance.
problem Pruning recurrent neural networks (RNNs) is challenging and often leads to poor performance.
method Data-efficient pruning objective derived from the spectrum of the recurrent Jacobian.
result 95% sparse GRUs significantly improve on existing baselines.
This work presents deep asymmetric networks with a set of node-wise variant activation functions. The nodes' sensitivities are affected by activation function selections such that the nodes with smaller indices become increasingly more sensitive. As a result, features learned by the nodes are sorted by the node indices…
We propose a new formulation for pruning convolutional kernels in neural networks to enable efficient inference. We interleave greedy criteria-based pruning with fine-tuning by backpropagation - a computationally efficient procedure that maintains good generalization in the pruned network. We propose a new criterion ba…
Dynamic pruning during training reduces deep network complexity without significant accuracy loss.
problem High memory and computational requirements of deep networks during training and inference.
method Dynamic pruning of convolutional filters during training, using L1 normalization for optimization.
result L1 normalization-based pruning yields up to 50% reduction in filters with minimal accuracy loss.
Proposes a method to improve surrogate models by incorporating sensitivity information.
problem Pruned neural networks often fail to capture sensitivities and uncertainties of original models.
method Combines Interval Adjoint Significance Analysis and Sobolev Training to accurately model sensitivities.
result Pruned models based on the proposed method better match original sensitivities.
i-SpaSP prunes neural networks by identifying important groups of parameters, improving pruning efficiency.
problem Pruning neural networks to reduce computational cost and improve performance.
method i-SpaSP uses sparse signal recovery principles to iteratively identify and threshold important parameter groups.
result i-SpaSP achieves strong empirical results and theoretical convergence guarantees, improving pruning efficiency.
Pruning filters in neural networks improves performance and reduces costs.
problem Large neural networks have high inference costs and limited resource usage.
method Batch Bridgeout for efficient pruning of convolutional filters.
result Batch Bridgeout trained networks achieve higher accuracy across various pruning intensities.
Holistic Filter Pruning reduces DNN complexity efficiently.
problem Redundant parameters in deep neural networks.
method Holistic Filter Pruning (HFP) for efficient DNN training.
result Achieves state-of-the-art performance with 60% reduction in multiplications on ImageNet.
A new method prunes neural networks faster and more efficiently.
problem Reducing training time and memory usage for neural networks.
method Set-based Task-Adaptive Meta Pruning (STAMP) that meta-learns a pruning mask.
result Significantly improved compression rates and faster training speed.
A new pruning method finds sparse minimizers in flat regions of deep neural networks.
problem Finding sparse minimizers in flat regions of deep neural networks.
method Directional pruning using an ℓ1 proximal gradient algorithm. result Empirical results show promising sparsity (92%) and comparable minima to SGD.
This research investigates selectively pruning hyper and hypo neurons to improve neural network generalization.
problem Improving neural network generalization to unseen data.
method Investigates pruning hyper and hypo neurons selectively in fully connected layers of CNNs.
result Selective pruning of hyper and hypo neurons improves model performance on out-of-domain data.
This work extends neural networks to automatically select features by stochastically penalizing feature involvement.
problem Feature selection in machine learning models.
method Stochastic regularization to select features instead of layer weights.
result Superior efficiency compared to classical methods with minimal computational overhead.
We show implicit filter level sparsity manifests in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization or weight decay. Through an extensive empirical study (Mehta et al., 2019) we hypothesize the mec…
To address the limitations of existing magnitude-based pruning algorithms in cases where model weights or activations are of large and similar magnitude, we propose a novel perspective to discover parameter redundancy among channels and accelerate deep CNNs via channel pruning. Precisely, we argue that channels reveali…
Expanding neural networks improves their learning from noisy data.
problem Improving neural network performance in noisy conditions.
method Sparse expansion of neural network inputs, followed by pruning, enhances generalization.
result Sparse expansion of neural networks improves generalization performance, even after pruning.
The study examines if ReLU activation function is optimal for modularity in neural networks.
problem Finding the best activation function for modularity in neural networks.
method Comparing ReLU with other activation functions for modularity and performance.
result ReLU may not be the best choice for modularity, suggesting other functions could be more suitable.
This paper introduces GLT for better input data representation in BNN and proposes a compact topology with block pruning.
problem Improving input data representation for Binary Neural Networks (BNN).
method Generic Learned Thermometer (GLT) for encoding, block pruning and Knowledge Distillation for compact topology.
result Significant accuracy gains and lightweight fully-binarized models with limited accuracy degradation.
Network pruning is aimed at imposing sparsity in a neural network architecture by increasing the portion of zero-valued weights for reducing its size regarding energy-efficiency consideration and increasing evaluation speed. In most of the conducted research efforts, the sparsity is enforced for network pruning without…
EVGAE improves VGAE's latent representation learning by mitigating over-pruning.
problem Over-pruning in VGAE limits latent variable capacity and diversity.
method EVGAE uses epitomic approach with multiple sparse VGAE models (epitomes) to increase active latent units and improve generative ability.
result EVGAE outperforms VGAE in generative ability and link prediction on citation networks.
PLOF reduces LOF computation time by pruning unnecessary data points.
problem Computational inefficiency of LOF in outlier detection.
method Prune-based approach to reduce LOF computation time.
result PLOF achieves higher accuracy and precision than original LOF and state-of-the-art variants.
Binary Stochastic Filtering (BSF), the algorithm for feature selection and neuron pruning is proposed in this work. The method defines filtering layer which penalizes amount of the information involved in the training process. This information could be the input data or output of the previous layer, which directly lead…
Neural networks are easier to optimise when they have many more weights than are required for modelling the mapping from inputs to outputs. This suggests a two-stage learning procedure that first learns a large net and then prunes away connections or hidden units. But standard training does not necessarily encourage ne…
Bayesian Bits unifies quantization and pruning through gradient optimization.
problem Joint mixed precision quantization and pruning for efficient neural networks.
method Gradient-based optimization with a novel bit width decomposition and learnable stochastic gates.
result Bayesian Bits achieves better accuracy vs. efficiency trade-off compared to static bit width networks.
Pruning CNNs by removing less important filters based on empirical loss changes.
problem Reducing memory and computation requirements for CNNs on resource-limited devices.
method Developed a novel filter importance norm based on empirical loss changes, and used sampling and ranking to prune filters.
result Reduced 60% of parameters and 64% of FLOPs with less than 0.6% accuracy drop.
SeReNe prunes neurons with low sensitivity to reduce network size.
problem Large neural networks consume too many resources on resource-constrained devices.
method Exploits neural sensitivity as a regularizer to prune neurons with low sensitivity.
result Pruning neurons with low sensitivity achieves competitive compression ratios.
New SDP method certifies neural network robustness across all classes efficiently.
problem Certifying robustness of neural networks across multiple classes.
method Quadratic model + SDP relaxation + pruning strategy.
result Significant computational speed-up and scalability to large datasets.
Bayesian neural networks improve macroeconomic forecasting and model nonlinearities.
problem Handling small T, big K macroeconomic datasets with temporal dependence.
method Developed Bayesian neural networks with mixture activation functions, shrinkage priors, and stochastic volatility.
result BNNs produce precise density forecasts, often better than other methods.
FS&P uses birth-death process to ensure global convergence of stochastic conic particle gradient descent.
problem Global optimization of non-convex objective functions over measure space.
method Introduces Fast Spawn\&Prune (FS\&P) combining CPGD with birth-death process.
result First theoretical guarantee of global convergence for discrete-time stochastic algorithms.
We present novel understandings of the Gamma-Poisson (GaP) model, a probabilistic matrix factorization model for count data. We show that GaP can be rewritten free of the score/activation matrix. This gives us new insights about the estimation of the topic/dictionary matrix by maximum marginal likelihood estimation. In…
Neural plasticity is an important functionality of human brain, in which number of neurons and synapses can shrink or expand in response to stimuli throughout the span of life. We model this dynamic learning process as an L0-norm regularized binary optimization problem, in which each unit of a neural network (e.g., …
Enhanced network threat detection using KG, LLM, and imbalanced learning.
problem Challenges in network threat detection due to complex attack patterns and limited historical data.
method Integrated framework combining Knowledge Graph, Imbalanced Learning, and Large Language Model.
result Improved threat capture rate by 3%-4% and increased interpretability of risk predictions.
Lookahead pruning extends single-layer optimization to multi-layer, outperforming magnitude-based pruning.
problem Pruning neural networks to reduce computational cost and memory usage.
method Developed a multi-layer optimization approach extending the single-layer optimization of magnitude-based pruning.
result Consistently outperforms magnitude-based pruning on various networks, especially in high sparsity.
Paper tackles active labeling for partial supervision.
problem Accessing stochastic gradients with partial supervision.
method Streaming technique to minimize generalization error.
result Proves minimization of generalization error ratio.