This note explains when neural networks can be seen as Gaussian processes.
problem Understanding the relationship between neural networks and Gaussian processes.
method Formulating a Gaussian process regression based on neural network outputs and analyzing the resulting posterior mean functions.
result The posterior mean functions of neural networks follow a Gaussian process in certain cases, providing an interpretation of reproducing kernel Hilbert spaces.
We propose a simple method that combines neural networks and Gaussian processes. The proposed method can estimate the uncertainty of outputs and flexibly adjust target functions where training data exist, which are advantages of Gaussian processes. The proposed method can also achieve high generalization performance fo…
Deep neural networks and Gaussian processes are shown to be equivalent through activation functions.
problem Understanding the relationship between neural networks and Gaussian processes.
method Developing an equivalence theory based on activation functions and kernels.
result Models can be seen as neural networks with improved uncertainty prediction or deep Gaussian processes with increased accuracy.
This work approximates finite neural networks with Gaussian processes, providing error bounds and applications in prior selection.
problem Approximating finite neural networks with Gaussian processes for error bounds and uncertainty quantification.
method Iterative approximation of neural network layers as mixtures of Gaussian processes, using optimal transport and Gaussian processes.
result The ability to return a mixture of Gaussian processes that is ε-close to the neural network at a finite set of input points.
New method for Bayesian neural networks with unbounded weights.
problem Posterior inference for Bayesian neural networks with unbounded weights.
method Conditionally Gaussian representation for efficient posterior inference.
result Interpretable and computationally efficient procedure for posterior inference.
The paper studies deep neural networks with Gaussian weights and finds their asymptotic behavior.
problem Understanding the behavior of deep neural networks with large width.
method Function-space perspective, Gaussian process analysis, weak convergence in large-width limit.
result Deep neural networks with large width converge to a continuous Gaussian process.
Study Gaussian-process limits of neural networks using tensor programs.
problem Understanding the behavior of neural networks as they approach infinite width.
method Quantitative analysis through tensor programs and Wasserstein distance.
result Explicit finite-width error bounds, showing convergence to Gaussian-process limits.
This paper shows how infinitely wide Tensor Networks converge to Gaussian Processes.
problem Understanding the relationship between Tensor Networks and Gaussian Processes.
method Analyzing the infinite-width limit of Tensor Networks and comparing them to Gaussian Processes.
result Infinitely wide Tensor Networks converge to Gaussian Processes, proving their equivalence.
Proposes scale mixture of NNGPs for more flexible stochastic processes.
problem Limited focus on broadening the class of stochastic processes from NNGPs.
method Scale mixture of NNGPs with scale priors on last-layer parameters.
result Turns neural networks into a richer class of stochastic processes.
Wide neural networks converge to Gaussian processes, improving generalization.
problem Understanding the generalization of wide neural networks, especially deep equilibrium models.
method Investigation of deep equilibrium models (DEQs) with infinite-depth layers, focusing on their convergence to Gaussian processes as width and depth approach infinity.
result Wide DEQs converge to Gaussian processes, maintaining generalization performance.
Study deep maxout networks and their equivalence to Gaussian processes.
problem Understanding neural networks with infinite width.
method Derive equivalence between deep maxout networks and Gaussian processes, characterize maxout kernel, and provide efficient numerical implementation.
result Bayesian inference based on deep maxout network kernel leads to competitive results compared to finite-width counterparts and deep neural network kernels.
Bayesian inference for wide neural networks using Edgeworth expansion.
problem Analyzing the non-Gaussian behavior of wide neural networks in Bayesian inference.
method Proposed a non-Gaussian distribution using multivariate Edgeworth expansion for finite-width neural networks.
result Derived non-Gaussian posterior distribution in Bayesian regression tasks.
Gaussian processes are ubiquitous in nature and engineering. A case in point is a class of neural networks in the infinite-width limit, whose priors correspond to Gaussian processes. Here we perturbatively extend this correspondence to finite-width neural networks, yielding non-Gaussian processes as priors. The methodo…
Bayesian neural networks use ridgelet prior for uncertainty quantification.
problem Combining strong predictive performance with uncertainty quantification in Bayesian neural networks.
method Proposes a ridgelet prior that approximates a Gaussian process covariance function in the output space of the network.
result Establishes universality property allowing Bayesian neural networks to approximate any Gaussian process.
Study shows neural networks trained with GD converge to Gaussian processes with polynomial decay.
problem Understanding convergence of neural networks to Gaussian processes during training.
method Explicit upper bounds on quadratic Wasserstein distance between trained networks and Gaussian approximations.
result Polynomial decay of approximation error with network width and training time.
Gaussian process models simplify neural network behavior for easier understanding.
problem Understanding and predicting the behavior of deep learning systems.
method Constructing surrogate models using Gaussian processes from finite neural networks.
result Surrogate models capture phenomena like spectral bias and predict generalization well.
Random neural networks with ReLU activations are non-Gaussian processes.
problem Understanding the behavior of neural networks with random initialization and rectified linear units.
method Proving these networks are non-Gaussian processes and deriving their properties.
result These networks can converge to non-Gaussian processes under certain conditions.
Wide neural networks can be closely approximated by Gaussian processes, with rates depending on the activation function's properties.
problem Approximating the behavior of wide neural networks using Gaussian processes.
method Established convergence rates for the central limit theorem in an infinite-dimensional functional space, using a transportation distance metric.
result Explicit convergence rates for neural networks approximated by Gaussian processes, varying based on the activation function's properties.
Study on MC dropout in wide neural networks and its convergence to Gaussian processes.
problem Understanding the behavior of Monte Carlo dropout in wide neural networks.
method Rigorously studied the limiting distribution of wide untrained NNs under dropout, proving convergence to Gaussian processes. Investigated correlations and non-Gaussian behavior in finite width NNs.
result Wide untrained neural networks under dropout converge to Gaussian processes for fixed sets of weights and biases.
Wide deep neural networks with Gaussian weights approximate Gaussian processes closely.
problem Understanding the approximation of deep neural networks with Gaussian weights to Gaussian processes.
method Established novel rates for the Gaussian approximation of random deep neural networks with Gaussian parameters and Lipschitz activation functions in the wide limit.
result The distance between the network output and the Gaussian approximation scales inversely with the width of the network.
Study of deep linear neural networks with proportional width and depth.
problem Lack of descriptive power in Gaussian limit of deep linear neural networks.
method Proportional infinite-width infinite-depth limit for deep linear neural networks.
result Characterization of limiting distribution as a nontrivial mixture of Gaussians.
The abstract proposes a neural network theory using quantum field theory.
problem Understanding the behavior of neural networks in the asymptotic and non-asymptotic limits.
method Mapping neural networks to Wilsonian effective field theory, using Gaussian processes and Feynman diagrams.
result Established a direct connection between overparameterization and simplicity of neural network likelihoods.
Study how depth affects inference in deep Bayesian neural networks.
problem Understanding how depth impacts inference in overparameterized linear Bayesian neural networks.
method Interpreting finite deep linear Bayesian neural networks as scale mixtures of Gaussian process predictors.
result Advances analytical understanding of how depth affects inference in a simple class of Bayesian neural networks.
New method renormalizes neural network Gaussian processes to identify learnable vs. unlearnable modes.
problem Separating learnable from unlearnable information in neural networks.
method Wilsonian renormalization applied to Gaussian Process Regression.
result Obtains a universal flow of the ridge parameter that becomes input-dependent.
In this paper we cast the well-known convolutional neural network in a Gaussian process perspective. In this way we hope to gain additional insights into the performance of convolutional networks, in particular understand under what circumstances they tend to perform well and what assumptions are implicitly made in the…
Quantum neural networks converge to Gaussian processes as they grow.
problem Understanding the convergence of quantum neural networks to Gaussian processes.
method Analyzing Haar random unitary and orthogonal deep QNNs, considering input states, measurement observables, and non-independence of unitary matrix entries.
result Quantum neural networks outputs converge to Gaussian processes in the limit of large Hilbert space dimension.
Bayesian optimization with Gaussian process as surrogate model has been successfully applied to analog circuit synthesis. In the traditional Gaussian process regression model, the kernel functions are defined explicitly. The computational complexity of training is O(N 3 ), and the computation complexity of prediction i…
Bounds neural network output distribution to Gaussian for random initialization.
problem Quantifying the distribution of randomly initialized deep neural networks.
method Quantitative Gaussian approximation using quadratic Wasserstein distance.
result Explicit inequalities show how network sizes affect Gaussian behavior.
Hybrid Bayesian neural networks use function uncertainty for probabilistic inference.
problem Uncertainty in neural network weights is hard to specify and interpret.
method Integrates probabilistic layers with standard deterministic layers for function uncertainty.
result Improves probabilistic inference by encoding function uncertainty.
Paper connects neural networks to Gaussian processes for understanding double-descent.
problem Understanding the double-descent phenomenon in neural networks.
method Uses techniques from random matrix theory and Gaussian processes.
result Establishes a connection between NNGP and random matrix theory for neural networks.
Wide neural networks can degrade performance, contrary to conventional wisdom.
problem Understanding the limitations of increasing network width in neural networks.
method Using Deep Gaussian Processes to decouple capacity and width, analyzing their effects on representational power and non-Gaussianity.
result Wide neural networks can become less adaptable and more Gaussian, leading to performance degradation.
Whilst deep neural networks have shown great empirical success, there is still much work to be done to understand their theoretical properties. In this paper, we study the relationship between random, wide, fully connected, feedforward networks with more than one hidden layer and Gaussian processes with a recursive ker…
Bayesian neural networks approximate Student-t processes in the infinite-width limit.
problem Modeling uncertainty in neural networks with greater flexibility.
method Extending asymptotic properties of Gaussian processes to Student-t processes in the infinite-width limit of BNNs.
result Posterior BNNs converge to Student-t processes in the infinite-width limit.
Quantitative CLTs show neural network distributions converge to Gaussian as width increases.
problem Understanding the distribution of fully connected neural networks with random weights and biases.
method Analyzing the distribution of a fully connected neural network with random Gaussian weights and biases, proving quantitative bounds on normal approximations.
result The distance between a random fully connected network and the corresponding infinite width Gaussian process scales like n−γ for γ>0. The neural tangent kernel equivalence theorem fails in practice.
problem Does the neural tangent kernel (NTK) equivalence theorem hold in practical neural network training?
method Rigorously derived NTK and conducted numerical experiments to evaluate the equivalence theorem.
result Adding a layer to a neural network and the corresponding updated NTK do not yield matching changes in predictor error.
We construct flexible likelihoods for multi-output Gaussian process models that leverage neural networks as components. We make use of sparse variational inference methods to enable scalable approximate inference for the resulting class of models. An attractive feature of these models is that they can admit analytic pr…
New approach to neural networks by incorporating observation noise and arbitrary prior means.
problem Misspecification on noisy data and limitations of NTK-GP equivalence.
method Introducing a regularizer for observation noise and proposing a shifted network for arbitrary prior means.
result Removes key obstacles to practical Gaussian process modeling in neural networks.
Post-hoc uncertainty quantification improves on pre-trained neural networks without underfitting.
problem Uncertainty quantification in neural networks is underfitting or computationally demanding.
method Gaussian Process Activation function (GAPA) for neuron-level uncertainty, with two methods: GAPA-Free and GAPA-Variational.
result GAPA-Variational outperforms Laplace approximation on most datasets in uncertainty quantification metrics.
Study of deep neural networks with dependent weights leading to new model limits and properties.
problem Characterizing deep neural networks with dependent weights in the infinite-width limit.
method Modeling weights as a mixture of Gaussian distributions and analyzing the infinite-width limit.
result Characterization of neural network layers by scalar parameters and Lévy measures, leading to new model limits.
GNet uses Gaussian processes for scalable, flexible neural networks.
problem Large-scale predictive modeling with high computational and storage costs.
method GNet employs Gaussian processes with nonparametric activation functions and a fast algorithm for training and predictions.
result GNet achieves competitive performance across various test problems, including nonlinear function prediction and real-world data regression.
GNet uses Gaussian processes for scalable, flexible neural networks.
problem Large-scale predictive modeling with high computational and storage costs.
method GNet employs Gaussian processes with nonparametric activation functions and a fast algorithm for efficient training and predictions.
result GNet achieves competitive performance across various test problems, including nonlinear function prediction and real-world data regression.
This paper extends the Gaussian process interpretation of deep networks to more varied weight distributions.
problem Understanding the impact of different weight initialization schemes on deep learning dynamics.
method Extending the Gaussian process interpretation to PSEUDO-IID weight distributions, including sparse and low-rank networks.
result PSEUDO-IID initialized networks are effectively equivalent up to variance, enabling tractable posterior distributions.
GWI combines deep neural networks with Gaussian processes for better predictive performance and uncertainty quantification.
problem Combining deep learning with Gaussian process uncertainty quantification.
method Gaussian Wasserstein inference (GWI) using Wasserstein distance between Gaussian measures.
result GWI achieves state-of-the-art performance on benchmark datasets.
NDPs learn to sample from complex function distributions using neural networks and diffusion models.
problem Learning rich distributions over functions with neural networks.
method NDPs use denoising diffusion models and custom attention blocks to incorporate stochastic process properties.
result NDPs can capture functional distributions close to true Bayesian posteriors and outperform neural processes.
Recent work has established the equivalence between deep neural networks and Gaussian processes (GPs), resulting in so-called neural network Gaussian processes (NNGPs). The behaviour of these models depends on the initialisation of the corresponding network. In this work, we consider the impact of noise regularisation …
Global inducing points improve Bayesian neural network performance.
problem Improving Bayesian neural network performance.
method Adapting correlated approximate posterior to all layers in a Bayesian neural network and deep Gaussian processes using learned global inducing points.
result State-of-the-art performance on CIFAR-10 (86.7%) without data augmentation or tempering.
A new method combines deep kernels with Gaussian processes to avoid overfitting.
problem Losing Bayesian benefits in deep kernel learning due to kernel optimization.
method Using Infinite-width neural networks and Neural Network Gaussian Process (NNGP) as a guide for DKL optimization.
result Robustness to overfitting and good predictive performance on various datasets.
Flexible nonstationary Gaussian process with neural network parameters.
problem Limited expressiveness of stationary Gaussian processes.
method Nonstationary kernels with neural network parameters trained jointly.
result Better accuracy and log-score compared to stationary and hierarchical models.