Study finds differences in LTs across tasks and architectures, proposing a consensus-based method for generating refined lottery tickets.
problem Understanding the variability and uniqueness of Lottery Tickets across different image classification tasks and architectures.
method 28 combinations of image classification tasks and architectures, iterative pruning techniques, consensus-based method for generating refined lottery tickets.
result Disproves the uniqueness of Lottery Tickets and connects emergent mask structure to the choice of pruning.
We find faster-converging sub-networks that significantly reduce adversarial training time.
problem Finding optimal sub-networks for adversarial training is costly and time-consuming.
method We identify a subset of sub-networks that converge faster during training.
result Sub-networks can reduce adversarial training time by up to 49%.
Early neural network training reveals important sub-networks and weight distributions.
problem Understanding the early phases of neural network training.
method Extensive measurements and quantitative probing of weight distribution and dataset reliance.
result Deep networks are not robust to reinitializing with random weights while maintaining signs, and weight distributions are highly non-independent.
We examine how recently documented, fundamental phenomena in deep learning models subject to pruning are affected by changes in the pruning procedure. Specifically, we analyze differences in the connectivity structure and learning dynamics of pruned models found through a set of common iterative pruning techniques, to …
The lottery ticket hypothesis finds multiple winning sub-networks in neural networks.
problem Finding a single winning sub-network in neural networks.
method Analyzing neural networks trained in isolation and on different tasks.
result Neural networks contain multiple sub-networks that match the accuracy of the original network, not just one.
New methods improve Laplace approximations for deep neural networks by selecting key parameters.
problem Improving uncertainty quantification in deep neural networks using computationally feasible approximations.
method Gradient-Laplace and Greedy-Laplace methods for selecting parameters in sub-network Laplace approximations.
result Gradient-Laplace method outperforms existing heuristic approaches and provides formal optimality guarantees.
Simplifies neural regression by combining two sub-networks for predictions and uncertainties.
problem Neural networks underestimate uncertainty, leading to overly confident predictions.
method Extends IRLS to a two-sub-network approach with shared representations and complementary loss functions.
result Proposed network is simpler to implement and more robust to uncertainty variations.
The pioneer deep neural networks (DNNs) have emerged to be deeper or wider for improving their accuracy in various applications of artificial intelligence. However, DNNs are often too heavy to deploy in practice, and it is often required to control their architectures dynamically given computing resource budget, i.e., …
AlphaNet improves supernets training with alpha-divergence.
problem Improving the uncertainty distillation in weight-sharing NAS.
method Proposes alpha-divergence for better uncertainty distillation in supernets.
result Significant improvements in model performance across various FLOPs regimes.
This paper considers a Bayesian view for estimating a sub-network in a Markov random field. The sub-network corresponds to the Markov blanket of a set of query variables, where the set of potential neighbours here is big. We factorize the posterior such that the Markov blanket is conditionally independent of the networ…
DC-NAS improves neural architecture search by clustering and evaluating sub-networks.
problem Inaccurate evaluation of neural architectures in large search spaces.
method Divide-and-Conquer approach: feature representation, clustering, and evaluation of clusters.
result Achieved 75.1% top-1 accuracy on ImageNet, surpassing state-of-the-art methods.
The recently proposed Lottery Ticket Hypothesis of Frankle and Carbin (2019) suggests that the performance of over-parameterized deep networks is due to the random initialization seeding the network with a small fraction of favorable weights. These weights retain their dominant status throughout training -- in a very r…
The capacity of a neural network to absorb information is limited by its number of parameters. Conditional computation, where parts of the network are active on a per-example basis, has been proposed in theory as a way of dramatically increasing model capacity without a proportional increase in computation. In practice…
The paper analyzes the role of ReLU gates in deep learning networks.
problem Understanding the role of gates in deep learning networks.
method Developed neural path features (NPF) and neural path values (NPV) to characterize the active sub-networks during training.
result The neural path kernel associated with NPFs is a fundamental quantity that characterizes the information stored in the gates of a DNN.
A distributed SGD method for heterogeneous networks with hubs and workers.
problem Learning in heterogeneous multi-level networks with worker heterogeneity and varying communication.
method Multi-Level Local SGD: distributed SGD with hub-and-spoke paradigm and hub averaging.
result The method converges with error dependent on worker heterogeneity, hub network topology, and iterations.
Pruning FCNs reveals sub-networks that match CNNs' performance.
problem Understanding the inductive bias of pruning in neural networks.
method Iterative magnitude pruning of a simple FCN followed by analysis of the resulting architecture.
result Pruned FCNs exhibit key features of CNNs, suggesting new architectural biases.
Adapts CNN for robust medical image segmentation across different scanners and protocols.
problem Performance degradation of CNNs in medical image segmentation due to mismatch between training and test images.
method Designs a segmentation CNN as a concatenation of a shallow normalization CNN and a deep CNN. At test time, adapts the normalization sub-network for each test image using a denoising autoencoder.
result Consistently improves performance on multi-center MRI datasets of brain, heart, and prostate.
In this paper, we study the trainability of rectified linear unit (ReLU) networks. A ReLU neuron is said to be dead if it only outputs a constant for any input. Two death states of neurons are introduced; tentative and permanent death. A network is then said to be trainable if the number of permanently dead neurons is …
Efficient trainable front-end for neural speech enhancement.
problem Inefficient STFT front-ends in neural speech enhancement models.
method Butterfly mechanism for Fast Fourier Transform, trainable STFT window.
result Accuracy and efficiency improvements for low-compute systems.
Study reveals trade dynamics in dry bulk shipping networks, highlighting their randomness and periodic changes.
problem Understanding the randomness and periodic changes in dry bulk shipping networks.
method Analysis of micro-level trade flow data from 2015 to 2023, focusing on grain, coal, and iron ore networks.
result Dry bulk shipping networks exhibit small-world phenomena and periodic life cycles, influenced by importing ports and global events.
SNAM improves NAM's accuracy and feature selection via group sparsity.
problem Improving interpretability and accuracy in deep learning models.
method Employing group sparsity regularization in neural additive models (SNAM).
result SNAM provably converges to zero training loss and achieves exact support recovery.
This paper uses graph convolutional networks to improve the accuracy of neural architecture search.
problem Improving the precision of sampled sub-networks in weight-sharing NAS.
method Training a graph convolutional network to fit the performance of sampled sub-networks.
result Achieved higher rank correlation coefficient and better final architecture performance.
Survey of trainable activation functions in neural networks.
problem Improving neural network performance through trainable activation functions.
method Taxonomy and comparison of recent and past models of trainable activation functions.
result Many trainable activation functions are equivalent to adding neuron layers with fixed activation functions and simple constraints.
Lottery tickets find good initializations for IMP with sparse training.
problem Finding good initializations for iterative magnitude pruning (IMP) in sparse networks.
method Empirical study of IMP performance with varying pre-training data and iterations.
result Training on a small fraction of data suffices to obtain good initializations for IMP.
Deep reinforcement learning for high dimensional, hierarchical control tasks usually requires the use of complex neural networks as functional approximators, which can lead to inefficiency, instability and even divergence in the training process. Here, we introduce stacked deep Q learning (SDQL), a flexible modularized…
This work explores the relation between trainability and dequantization in variational QML models.
problem Understanding the interplay between trainability and dequantization in variational QML models.
method Provide precise definitions of trainability and dequantization, study their relation, and introduce recipes for building PQC-based QML models.
result Identify conditions under which trainability and non-dequantization are not mutually exclusive.
iGNN tackles inverse graph prediction using invertible neural networks.
problem Inverse graph prediction problem in data analysis and machine learning.
method Developed invertible graph neural network (iGNN) to solve inverse prediction problem on graphs.
result iGNN model allows efficient generation from output labels and forward prediction.
Bayesian principles improve neural additive models for better feature selection and uncertainty.
problem Lack of calibrated uncertainties and feature selection in neural additive models.
method Augmenting NAMs with Bayesian principles to provide credible intervals, feature selection, and interaction ranking.
result Improved performance on tabular datasets and real-world medical tasks.
New framework improves interpretability of trainable prompts.
problem Improving task-specific LLM performance with soft prompts remains a black-box method.
method Developed a theoretical framework for evaluating interpretability of trainable prompts, inspired new objective functions.
result Found a fundamental trade-off between interpretability and task performance in trainable prompts.
We consider a general framework for reducing the number of trainable model parameters in deep learning networks by decomposing linear operators as a product of sums of simpler linear operators. Recently proposed deep learning architectures such as CNN, KFC, Dilated CNN, etc. are all subsumed in this framework and we il…
New proof links initial class bias to DNN trainability, challenging traditional understanding.
problem Understanding the initial class bias in DNNs and its impact on trainability.
method Theoretical proof linking initial class bias to mean field theories of DNNs.
result Efficient learning is connected to a network's prejudice towards a specific class, contradicting traditional understanding.
Predicts trainability of deep neural networks using reconstruction entropy.
problem Predicting the initial conditions for trainability of deep neural networks.
method Cascade of auxiliary networks to reconstruct input from activation layers, computing relative entropy.
result Predicts trainability of deep feedforward networks on various datasets with a single epoch.
In this paper, we present UNet++, a new, more powerful architecture for medical image segmentation. Our architecture is essentially a deeply-supervised encoder-decoder network where the encoder and decoder sub-networks are connected through a series of nested, dense skip pathways. The re-designed skip pathways aim at r…
Architecture optimization, which is a technique for finding an efficient neural network that meets certain requirements, generally reduces to a set of multiple-choice selection problems among alternative sub-structures or parameters. The discrete nature of the selection problem, however, makes this optimization difficu…
PatchGT uses non-trainable graph patches to improve graph representation learning.
problem Learning high-level information in graph tasks with direct Transformer models.
method PatchGT segments graphs into non-trainable patches, uses GNN for patch-level learning, and Transformer for graph-level learning.
result PatchGT achieves higher expressiveness and competitive performance on benchmark datasets.
A new method clusters subjects based on brain networks without vectorizing fMRI data.
problem Distortion of clustering results when simplifying fMRI data structure.
method Wishart mixture models for multiple-view clustering of brain networks.
result Identifies multiple underlying pairs of associations between subject clusters and brain sub-networks.
Quantum machine learning faces challenges similar to variational quantum algorithms in training.
problem Challenges in training quantum machine learning models.
method Bridge between variational quantum algorithms and quantum machine learning, applying gradient scaling results.
result Gradient scaling results for variational quantum algorithms can also be applied to quantum machine learning models, revealing new trainability issues.
Quantum ML promises faster data analysis but faces trainability challenges.
problem Challenges in training quantum machine learning models.
method Review of current methods and applications of quantum neural networks and quantum deep learning.
result Opportunities for quantum advantage in quantum machine learning.
Deep learning has achieved impressive prediction accuracies in a variety of scientific and industrial domains. However, the nested non-linear feature of deep learning makes the learning highly non-transparent, i.e., it is still unknown how the learning coordinates a huge number of parameters to achieve a decision makin…
Modeling financial time series with LSTM and trainable initial states.
problem Extracting patterns and information from financial time series.
method Long Short-Term Memory (LSTM) network with trainable initial hidden states.
result Model captures relative similarity and predicts future stock trends.
DARTS is a popular algorithm for neural architecture search (NAS). Despite its great advantage in search efficiency, DARTS often suffers weak stability, which reflects in the large variation among individual trials as well as the sensitivity to the hyper-parameters of the search process. This paper owes such instabilit…
Online discussion forums are complex webs of overlapping subcommunities (macrolevel structure, across threads) in which users enact different roles depending on which subcommunity they are participating in within a particular time point (microlevel structure, within threads). This sub-network structure is implicit in m…
New method uses trainable activations to make BNNs behave like GPs.
problem Making Bayesian Neural Networks (BNNs) behave like Gaussian Processes (GPs).
method Introduced trainable activations and periodic activations to map GP priors to BNNs. Used 2-Wasserstein distance for optimization.
result Method consistently outperforms existing approaches or matches heuristic methods.
Investigates ways to train larger models with fewer resources, finding that test loss depends only on the actual number of trainable parameters.
problem Training larger models for cheaper under hardware constraints.
method Emulates an increase in effective parameters using frozen random parameters or fast structured transforms.
result Scaling laws cannot be deceived by spurious parameters; test loss depends only on the actual number of trainable parameters.
Quantum models face barren plateaus, but specific losses can be trainable.
problem Barren plateaus and loss concentration in quantum generative models.
method Investigated explicit and implicit losses, and their interplay.
result Explicit losses lead to new barren plateaus, while implicit losses can be trainable.
Different types of Convolutional Neural Networks (CNNs) have been applied to detect cancerous lung nodules from computed tomography (CT) scans. However, the size of a nodule is very diverse and can range anywhere between 3 and 30 millimeters. The high variation of nodule sizes makes classifying them a difficult and cha…
QCNNs avoid barren plateaus, making them trainable.
problem Exponentially vanishing gradients in QNNs.
method Graph-based method to analyze Haar-distributed unitaries.
result QCNNs do not exhibit barren plateaus, implying trainability.
A new training method improves stability and generalization of DeepONets.
problem Training deep operator networks (DeepONets) is challenging due to nonconvex and nonlinear nature.
method Two-step training method: first train trunk network, then branch network. Introduced Gram-Schmidt orthonormalization.
result Generalization error estimate and numerical examples demonstrating effectiveness.