New neural network approach using mutual information.
problem Training neural networks for imbalanced datasets.
method Converts neural network classifiers to mutual information evaluators.
result New form of softmax leads to better classification accuracy, especially for imbalanced datasets.
Measures neural network information transfer for generalization.
problem Estimating the generalizable information in neural networks.
method Proposes Information Transfer (LIT) based on prequential coding. result Consistently correlates with generalizable information in neural networks.
Proposes a new method to enhance neural learning by maximizing information gain.
problem Improving neural learning by selecting key variables to maximize information gain.
method Adaptive Ensemble Kalman Filter to quantify uncertainty and maximize information gain.
result The proposed method enables the neural network to learn more effectively from stochastic systems.
Review of information plane analyses in neural networks, highlighting mixed results and methodological challenges.
problem Understanding the relationship between information-theoretic compression and neural network performance.
method Literature review and detailed analysis of information quantity estimation methods.
result Information plane compression is not necessarily information-theoretic but compatible with geometric compression.
Despite great popularity of applying softmax to map the non-normalised outputs of a neural network to a probability distribution over predicting classes, this normalised exponential transformation still seems to be artificial. A theoretic framework that incorporates softmax as an intrinsic component is still lacking. I…
Theory proposes neural networks can be initialized for optimal information transmission.
problem Optimizing neural networks for optimal information transmission and representation.
method Developed a corrected mean-field framework to study neural networks as information channels, proving mutual information maximization at dynamic isometry.
result Mutual information maximization is realized between inputs and propagated signals when neural networks are initialized at dynamic isometry.
A measure of neural complexity quantifies how hard it is to access information across neurons.
problem Understanding how mutual information is distributed among neurons in neural networks.
method Partial Information Decomposition (PID) to disentangle contributions of single neurons, multiple neurons, and synergistic effects.
result Representational Complexity measures the difficulty of accessing information across multiple neurons.
nnLDA combines neural and probabilistic methods for better topic modeling with side information.
problem Lack of integration of auxiliary information in traditional topic models.
method nnLDA integrates side information through a neural prior mechanism, optimizing both neural and probabilistic components.
result nnLDA outperforms traditional models in topic coherence, perplexity, and classification.
We find the maximum mutual information for neural networks and its key determinants.
problem Understanding the maximum mutual information in neural architectures.
method Derived closed-form expression for maximum mutual information across neural network families.
result Maximum mutual information stems from a generalized formula and is influenced by network width and statistical invariances.
The study examines Fisher information matrices and neural tangent kernels for simple ReLU networks with random weights.
problem Understanding the relationship between Fisher information matrices and neural tangent kernels for 2-layer ReLU networks.
method Analyzes Fisher information matrices and neural tangent kernels for 2-layer ReLU networks with random hidden weights, focusing on spectral decomposition and eigenfunctions.
result Obtained an approximation formula for functions represented by 2-layer neural networks.
Unified information-theoretic objectives for training deep neural networks.
problem Difficulty in computing information-theoretic quantities for large deep neural networks.
method Review and unify competing objectives, develop surrogate objectives.
result Surrogate objectives allow applying information bottleneck to modern neural network architectures.
Graph neural controlled differential equations learn graph dynamics from vertex observations.
problem Predicting future states of dynamical systems on graphs with limited vertex data.
method Incorporates graph topology information into NCDE to predict graph dynamics.
result Informed NCDE requires fewer parameters and lower MAE compared to previous methods.
Neural estimator improves mutual information estimation in high dimensions.
problem Estimating mutual information in high dimensions is challenging.
method Parametrizing conditional densities with normalizing flows and using block autoregressive structure.
result Improved mutual information estimation on benchmark tasks.
Study suggests using information flow measures to target interventions in neural networks.
problem Identifying neural network edges that can be pruned to reduce bias.
method Used M-information flow framework to measure and compare information flows about true labels and protected attributes, and evaluated pruning effects on bias reduction. result Pruning edges with larger information flows about protected attributes reduces bias at the output.
LogDet estimator improves entropy estimation in neural networks.
problem Inconsistent observations and diversified interpretation in neural networks.
method Proposes LogDet estimator for reliable entropy approximation.
result LogDet estimator overcomes distributional diversity issues.
New DDMs use neural networks for solving equations on manifold shapes.
problem Solving equations on complex, high-dimensional shapes.
method Physics-informed neural networks combined with domain decomposition methods.
result Validated methods work well on various shapes in high dimensions.
In this paper, we present a new approach to interpret deep learning models. By coupling mutual information with network science, we explore how information flows through feedforward networks. We show that efficiently approximating mutual information allows us to create an information measure that quantifies how much in…
Improved meta-learning for dynamics using additional structured knowledge.
problem Meta-learning for dynamics with limited raw observations.
method Extended Neural ODE Process model to use privileged information.
result Improved accuracy and calibration on simulated dynamics tasks.
Two Fisher information matrix estimators are analyzed for neural networks, focusing on their variances and trade-offs.
problem Estimating the Fisher information matrix in neural networks due to its high computational cost.
method Examined two popular diagonal Fisher information matrix estimators and their variances in neural networks for regression and classification.
result The variances of the estimators depend on the non-linearity with respect to different parameter groups and should not be neglected.
We argue that the estimation of mutual information between high dimensional continuous random variables can be achieved by gradient descent over neural networks. We present a Mutual Information Neural Estimator (MINE) that is linearly scalable in dimensionality as well as in sample size, trainable through back-prop, an…
Enhances neural networks with prior function values to improve accuracy.
problem Improving neural network accuracy in regions without training data.
method Develops a probabilistic approach to augment BNNs with prior function values.
result Predictions rely more on prior information in uncertain regions.
Physics-informed neural networks improve pathloss prediction accuracy.
problem Improving pathloss prediction accuracy in wireless communications.
method Physics-informed neural networks incorporating physical dependencies and measured values.
result Physics-informed neural networks achieve better generalization and prediction quality with fewer layers and parameters.
This study compares different thermodynamic structure-informed neural networks for solving differential equations.
problem Improving the accuracy and physical consistency of neural network solutions to differential equations.
method Comprehensive evaluation of various thermodynamic formulations in physics-informed neural networks.
result Newtonian-residual-based PINNs fail to reliably recover physical quantities, while structure-preserving formulations enhance accuracy and robustness.
New method solves PDEs on spheres using physics-informed convolutional neural networks.
problem Solving PDEs on surfaces, especially spheres, with high accuracy and efficiency.
method Physics-informed convolutional neural networks (PICNN) with theoretical analysis and approximation results.
result Established fast convergence rates for PICNN solving PDEs on spheres.
Study quantifies information flow in neural networks using relative entropy and RG analogy.
problem Quantifying information flow in deep neural networks.
method Explicit computation of relative entropy in Ising models and feedforward neural networks.
result Monotonic increase of relative entropy to an asymptotic value, confirming connection to c-theorem.
Bayesian neural networks improve with summary information and Dirichlet process.
problem Lack of prior knowledge in BNNs for complex architectures.
method Incorporates external summary information about predicted probabilities using a Dirichlet process.
result Improves model accuracy, uncertainty calibration, and robustness.
We extend manifold capacity to nonlinear neural representations with contextual information.
problem Efficient processing of information through neural representations.
method Theoretical framework leveraging latent directions in input space related to contextual information.
result Derivation of an exact formula for context-dependent manifold capacity.
New method estimates spin system mutual information using neural networks.
problem Estimating mutual information in spin systems.
method Monte Carlo sampling enhanced by autoregressive neural networks.
result Area law satisfied for temperatures away from critical temperature.
Whatever information a deep neural network has gleaned from training data is encoded in its weights. How this information affects the response of the network to future data remains largely an open question. Indeed, even defining and measuring information entails some subtleties, since a trained network is a determinist…
Study neural networks by mapping correlations, revealing essential statistics.
problem Understanding information processing in trained neural networks.
method Characterize neural network as distribution transformations, focusing on correlation functions.
result Higher-order correlations are crucial for internal layers, while input layer captures more.
This work integrates differentiation and integration in Physics-Informed Neural Networks.
problem Solving integro-differential equations and computing integral transforms.
method Augmenting Physics-Informed Neural Networks with automatic integration.
result Solving complex integral transforms and integro-differential equations.
Lifelong learning is a very important step toward realizing robust autonomous artificial agents. Neural networks are the main engine of deep learning, which is the current state-of-the-art technique in formulating adaptive artificial intelligent systems. However, neural networks suffer from catastrophic forgetting when…
Estimates how much samples inform neural network training and function.
problem Measuring informativeness of samples in neural networks.
method Linearized network approximations for efficient computation.
result Efficient approximations show good accuracy for real-world models.
Adaptive weights improve physics-informed neural networks and deep operator networks.
problem Training physics-informed neural networks and deep operator networks can be challenging, leading to unsatisfactory accuracy and efficiency.
method Proposes a pointwise adaptive weighting method that balances the residual decay rate across different training points.
result Our proposed approach of balanced residual decay rates offers advantages including bounded weights, high prediction accuracy, fast convergence rate, low training uncertainty, low computational cost, and ease of hyperparameter tuning.
Physics-informed neural networks simulate radiative transfer efficiently.
problem Simulating radiative transfer accurately and efficiently.
method Physics-informed neural networks trained to minimize radiative transfer equations.
result PINNs provide an easy-to-implement, robust, and accurate method for radiative transfer simulation.
This paper analyzes the training dynamics of binary neural networks using information bottleneck.
problem Training binary neural networks is challenging due to discontinuity in activation functions.
method The approach uses the Information Bottleneck principle to analyze BNN training dynamics.
result Training dynamics of BNNs are different from DNNs, with both phases occurring simultaneously.
jinns is a JAX library for physics-informed neural networks.
problem Physics-informed neural networks for forward and inverse problems.
method Physics-informed neural networks using JAX ecosystem.
result Efficient prototyping and extensions for real problems.
Neural networks have been proposed recently for positioning and channel charting of user equipments (UEs) in wireless systems. Both of these approaches process channel state information (CSI) that is acquired at a multi-antenna base-station in order to learn a function that maps CSI to location information. CSI-based p…
Improves inference from sparse data with hybrid summary statistics.
problem Robust simulation-based inference from limited data.
method Augment traditional summary statistics with neural network outputs to maximize mutual information.
result Improves information extraction and makes inference robust in low-data settings.
New bounds improve neural network generalization through slicing.
problem Difficulty in evaluating mutual information in high dimensions for neural networks.
method Slicing the parameter space and using disintegrated mutual information and k-sliced mutual information.
result Slicing improves generalization and offers significant computational and statistical advantages.
New method extracts cosmological information from dark matter halo catalogues using graph neural networks.
problem Quantifying cosmological information from large-scale structure data.
method Implicit likelihood approach with Information Maximising Neural Networks (IMNNs) on graph representations of dark matter halo catalogues.
result Graph neural network summaries can extract information from noisy catalogues and improve parameter constraints.
Graphs are essential representations of many real-world data such as social networks. Recent years have witnessed the increasing efforts made to extend the neural network models to graph-structured data. These methods, which are usually known as the graph neural networks, have been applied to advance many graphs relate…
Dropout technique is analyzed using information geometry.
problem Understanding the regularization performance of dropout in neural networks.
method Unified analysis from information geometry viewpoint.
result Dropout flattens the model manifold and its performance depends on curvature.
Enhances weak lensing inference with neural summaries.
problem Extracting additional information from weak lensing convergence maps.
method Hybrid approach combining physics-based and neural summaries.
result Neural summaries extract up to 8 times more information than angular power spectra.
New method decomposes sensory information from neurons into specific stimuli and features.
problem Understanding how much and what specific information neurons encode.
method Introduced axioms for meaningful stimulus-wise decomposition and derived a tractable solution using diffusion models.
result Can efficiently estimate contributions of specific stimuli and features to encoded information.
Physics-informed neural networks and neural operators speed up solving parametric PDEs by orders of magnitude.
problem Solving PDEs for varying parameters is computationally expensive.
method Physics-informed neural networks and neural operators learn solution mappings across parameter spaces.
result Neural operators achieve computational speedups of 10^3 to 10^5 times faster than traditional methods.
We develop a sparse representation method for neural network uncertainty.
problem Estimating model uncertainty in neural networks.
method Sparse representation of model uncertainty using inverse Multivariate Normal Distribution (MND), with a novel sparsification algorithm and analytical sampler.
result The information form of neural networks can be effectively applied for model uncertainty representation, showing competitive performance.
Unified framework for Bayesian PDE-constrained inversion using physics-informed neural networks.
problem Incorporating prior distributions in function space into Bayesian PINN-based inversion.
method Functional-prior-based approaches (fpBPINN) to Bayesian PDE-constrained inversion using physics-informed neural networks (PINNs). Two complementary approaches: FPI-BPINN and fParVI-PINN.
result Accurate estimation of posterior distributions in seismic traveltime tomography and Darcy-flow permeability inversion.