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.
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 …
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.
We propose reinforcement learning on simple networks consisting of random connections of spiking neurons (both recurrent and feed-forward) that can learn complex tasks with very little trainable parameters. Such sparse and randomly interconnected recurrent spiking networks exhibit highly non-linear dynamics that transf…
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.
New approach connects quantum phases to VQA trainability, enabling better scaling.
problem Scalability issues in VQAs, especially barren plateaus.
method Analog VQA ansätze composed of quenches of a disordered Ising chain, tuning disorder strength.
result Thermalized and MBL phases reach maximal expressivity at large M, but barren plateaus emerge at smaller M in the thermalized phase. Two-layer CNNs can overfit well if initialized correctly.
problem Understanding the conditions for benign overfitting in over-parameterized CNNs.
method Extending analysis to fully trainable two-layer CNNs, examining initialization scaling effects.
result Initialization scaling of the output layer is crucial; large scales lead to fixed output behavior, small scales to complex interactions.
New algorithm finds important synapses without training data.
problem Finding important synapses in neural networks without data.
method Iterative Synaptic Flow Pruning (SynFlow) based on conservation law.
result Algorithm consistently outperforms existing pruning methods.
New insights into neural network training efficiency.
problem Understanding the optimal initialization for deep neural networks.
method Exploring the edge of chaos and saturation of tanh activation function.
result The line of uniformity in phase space intersects the edge of chaos, indicating saturation begins to hinder training efficiency.
Barren plateaus are not an average-case phenomenon, but a highly non-unique problem.
problem Avoiding barren plateaus in neural network training
method First-moment framework for initialization strategies
result Many families of inequivalent initialization strategies can avoid concentration
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.
Echo State Networks (ESN) are a class of Recurrent Neural Networks (RNN) that has gained substantial popularity due to their effectiveness, ease of use and potential for compact hardware implementation. An ESN contains the three network layers input, reservoir and readout where the reservoir is the truly recurrent netw…
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.
Stability of recurrent models is closely linked with trainability, generalizability and in some applications, safety. Methods that train stable recurrent neural networks, however, do so at a significant cost to expressibility. We propose an implicit model structure that allows for a convex parametrization of stable mod…
Study of infinitely deep but narrow neural networks using NTK theory.
problem Analyzing the role of depth in deep learning with overparameterized networks.
method Infinite-depth limit analysis of MLP and CNN using Neural Tangent Kernel (NTK) theory.
result Established trainability guarantee for infinitely deep but narrow neural networks.
EVA adapts LoRA for faster, more efficient fine-tuning.
problem Fast and efficient fine-tuning of large models for specific tasks.
method EVA uses directions capturing most activation variance for initialization, maximizing gradient signal and reducing parameters.
result EVA achieves faster convergence and higher average scores across tasks, reducing parameters.
Paper improves deep learning for solving evolutionary equations with trainable hard constraints.
problem Low computational accuracy of standard PINNs in large temporal domains.
method Sequential learning strategies and trainable influence functions for hard constraints.
result Significantly improved computational accuracy and universality of the method.
Probabilistic bounds on neuron death in deep networks, showing depth can be increased indefinitely.
problem Understanding neuron death in deep rectifier networks and its impact on model trainability.
method Deriving upper and lower bounds on neuron death probability as a function of model hyperparameters.
result The probability of neuron death decreases as network depth increases, provided width increases proportionally.
Deep learning relies on good initialization schemes and hyperparameter choices prior to training a neural network. Random weight initializations induce random network ensembles, which give rise to the trainability, training speed, and sometimes also generalization ability of an instance. In addition, such ensembles pro…
New method removes contrastive loss by adding a prediction head, revealing learning mechanisms.
problem Understanding why neural networks learn competitive representations despite trivial optima.
method Empirical and theoretical analysis of a trainable, identity-initialized prediction head.
result The trainable prediction head enables learning all features, preventing dimensional collapse.
Recurrent neural networks have gained widespread use in modeling sequence data across various domains. While many successful recurrent architectures employ a notion of gating, the exact mechanism that enables such remarkable performance is not well understood. We develop a theory for signal propagation in recurrent net…
Quantum neural networks need both data-dependent and trainable unitaries for effective geometric deformation.
problem Quantum neural networks lack the geometric flexibility of classical networks due to limitations in state reachability.
method Viewing quantum states as embedded manifolds, we analyze infinitesimal unitary actions and introduce the CLA maps and aCLS criterion.
result Geometric flexibility in quantum neural networks requires a joint dependence on data and trainable weights.
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.
We introduce MinimalRNN, a new recurrent neural network architecture that achieves comparable performance as the popular gated RNNs with a simplified structure. It employs minimal updates within RNN, which not only leads to efficient learning and testing but more importantly better interpretability and trainability. We…
LSQ+ improves quantization of neural nets with Swish activations, achieving state-of-the-art results.
problem Quantization of neural nets with Swish activations, especially negative activations, leads to significant performance loss.
method Introduces learnable scale and offset parameters for asymmetric quantization, and uses MSE-based initialization for quantization parameters.
result Significantly outperforms LSQ for low-bit quantization of neural nets with Swish activations, achieving up to 5.6% gain with W2A2 quantization of EfficientNet-B0.
Warm starts improve variational quantum algorithms by avoiding barren plateaus.
problem Barren plateaus in variational quantum algorithms limit scaling.
method Exploring warm starts in iterative variational methods for quantum circuits.
result Warm starts can lead to substantial gradients in small regions, suggesting trainability.
Reducing the precision of weights and activation functions in neural network training, with minimal impact on performance, is essential for the deployment of these models in resource-constrained environments. We apply mean-field techniques to networks with quantized activations in order to evaluate the degree to which …
Neural networks benefit from intermediate representations, reducing sample complexity.
problem Understanding how neural networks leverage intermediate representations for hierarchical learning.
method Fixed, randomly initialized neural network as a representation function, compared with raw inputs and other trainable networks.
result Neural representations can achieve improved sample complexities compared to raw inputs, especially for low-rank polynomials.
Network pruning is a promising avenue for compressing deep neural networks. A typical approach to pruning starts by training a model and then removing redundant parameters while minimizing the impact on what is learned. Alternatively, a recent approach shows that pruning can be done at initialization prior to training,…
A quantum circuit designed for efficient statistical model preparation and training.
problem Challenges in preparing and learning statistical models on quantum processors.
method Utilizes the maximum entropy principle to design a statistics-informed parameterized quantum circuit (SI-PQC).
result Improves trainability and interpretability for learning quantum states and classical model parameters.
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.
Variational quantum computing faces a flat optimization landscape problem.
problem Barren Plateaus (BP) in optimization landscapes.
method Theoretical and heuristic methods to understand and mitigate BPs.
result All algorithm components can lead to BPs if not well-suited.
State construction is important for learning in partially observable environments. A general purpose strategy for state construction is to learn the state update using a Recurrent Neural Network (RNN), which updates the internal state using the current internal state and the most recent observation. This internal state…
Many neural speech enhancement and source separation systems operate in the time-frequency domain. Such models often benefit from making their Short-Time Fourier Transform (STFT) front-ends trainable. In current literature, these are implemented as large Discrete Fourier Transform matrices; which are prohibitively inef…
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.
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.
Study evaluates capacity and trainability of parametrized quantum circuits.
problem Finding the best type of circuits for hybrid quantum-classical algorithms.
method Geometric structure of parameter space, effective quantum dimension, and circuit expressiveness.
result Identifies a transition in quantum geometry leading to decay of quantum natural gradient for deep circuits.
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…
Due to the widespread use of positron emission tomography (PET) in clinical practice, the potential risk of PET-associated radiation dose to patients needs to be minimized. However, with the reduction in the radiation dose, the resultant images may suffer from noise and artifacts that compromise diagnostic performance.…
Survey on reducing deep neural network training complexity.
problem Efficient training of large deep neural networks.
method Pruning and freezing parts of the network during training.
result Dimensionality reduction improves training efficiency.
End-to-end trainable graph matching using improved combinatorial solvers.
problem Graph matching in deep learning.
method Combining deep learning with optimized combinatorial solvers.
result Advances state-of-the-art on deep graph matching benchmarks.
CLOPS improves deep learning for continuous physiological data.
problem Deep learning struggles with streaming, multi-sensor clinical data.
method CLOPS is a replay-based continual learning strategy.
result CLOPS outperforms state-of-the-art methods in continual learning scenarios.
Batch normalization dramatically increases the largest trainable depth of residual networks, and this benefit has been crucial to the empirical success of deep residual networks on a wide range of benchmarks. We show that this key benefit arises because, at initialization, batch normalization downscales the residual br…
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…
The weight initialization and the activation function of deep neural networks have a crucial impact on the performance of the training procedure. An inappropriate selection can lead to the loss of information of the input during forward propagation and the exponential vanishing/exploding of gradients during back-propag…
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.
APINNs improve physics-informed neural networks through flexible domain decomposition.
problem Improving physics-informed neural networks (PINNs) for solving partial differential equations (PDEs).
method Introduces a trainable gate network for soft domain decomposition, allowing flexible parameter sharing and improved generalization.
result APINNs significantly improve PINNs and XPINNs, demonstrating better performance on various types of PDEs.
Improved neural ODEs learn adaptable flows.
problem Neural ODEs struggle with expressive power and adaptability.
method Introduce N-CODE modules with dynamic parameters controlled by a trainable map.
result N-CODE modules enhance expressivity of neural ODEs.