Invertible networks help explain decisions and identify important features.
problem Interpreting and explaining the decisions of black-box neural networks.
method Two-stage approach: invertible transformation to feature space and linear classifier. Determining decision boundaries and feature importance using local linear models.
result Ability to explain decisions and identify important features in neural networks.
Residual neural networks don't help overcome sampling complexity issues.
problem Learning invertible residual neural networks from samples is hard due to the curse of dimensionality.
method Investigated invertible residual neural networks and their sampling complexity.
result Invertible residual neural networks still suffer from the curse of dimensionality in sampling complexity.
Proposes a constant memory iterative inverse model using invertible networks.
problem Memory limitations in iterative learning approaches for inverse problems.
method Invertible networks to avoid storing intermediate activations, constant memory model.
result Trains 400-layer models on 3D MRI volumes, achieving state-of-the-art image reconstruction.
We propose a new way of constructing invertible neural networks by combining simple building blocks with a novel set of composition rules. This leads to a rich set of invertible architectures, including those similar to ResNets. Inversion is achieved with a locally convergent iterative procedure that is parallelizable …
ISR creates analytical relationships from data via invertible maps.
problem Creating analytical relationships from datasets.
method Combines INNs and EQL, using invertible maps and sparsity promoting regularization.
result ISR can serve as a normalizing flow for density estimation and solve inverse problems.
Generative model for 3D point clouds using invertible flows.
problem Generating realistic 3D point clouds.
method Invertible flow-based models for point cloud generation with parameter sharing and embedding vectors.
result The model generates high-quality 3D point clouds with good similarity.
Flow-based generative models parameterize probability distributions through an invertible transformation and can be trained by maximum likelihood. Invertible residual networks provide a flexible family of transformations where only Lipschitz conditions rather than strict architectural constraints are needed for enforci…
Invertible DenseNets improve model efficiency and performance.
problem Improving model efficiency and performance in neural networks.
method Enforcing invertibility in DenseNets by satisfying the Lipschitz constraint and proposing a learnable concatenation.
result i-DenseNets outperform Residual Flows in negative log-likelihood on various datasets.
This work tackles exploding inverses in INNs, revealing and mitigating their numerical non-invertibility.
problem Exploding inverses in INNs cause numerical non-invertibility, leading to failures in various tasks.
method Derived bi-Lipschitz properties of INN building blocks, proposed regularizers for local invertibility, and stable INN designs for global invertibility.
result Bi-Lipschitz properties and stable INN designs are crucial for addressing numerical non-invertibility.
CF-INNs can approximate any invertible function, resolving a long-standing problem.
problem Whether CF-INNs can approximate any invertible function.
method Demonstrated CF-INNs are universal approximators for invertible functions by showing a convenient criterion.
result CF-INNs are universal approximators for invertible functions.
The paper stabilizes invertible neural networks by using Gaussian mixture models.
problem Invertible neural networks can have exploding Lipschitz constants, leading to numerical errors.
method The authors use Gaussian mixture models to stabilize the latent distribution of invertible neural networks.
result Numerical simulations confirm that this modification improves sampling quality in multimodal applications.
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.
We show that standard ResNet architectures can be made invertible, allowing the same model to be used for classification, density estimation, and generation. Typically, enforcing invertibility requires partitioning dimensions or restricting network architectures. In contrast, our approach only requires adding a simple …
Electronic power inverters are capable of quickly delivering reactive power to maintain customer voltages within operating tolerances and to reduce system losses in distribution grids. This paper proposes a systematic and data-driven approach to determine reactive power inverter output as a function of local measuremen…
Paper develops a method to learn causal networks with non-invertible functions.
problem Identifying causal relationships from observational data with non-invertible functional relationships.
method Proposes a test for non-invertible bivariate causal models and develops a method to incorporate this test in structure learning of DAGs.
result Our algorithms outperform existing DAG learning methods in identifying causal graphical structures.
Studying the invertibility of deep neural networks (DNNs) provides a principled approach to better understand the behavior of these powerful models. Despite being a promising diagnostic tool, a consistent theory on their invertibility is still lacking. We derive a theoretically motivated approach to explore the preimag…
This work presents a novel approach to train invertible linear layers by adding rank-one perturbations.
problem Training invertible linear layers during optimization with gradient-based methods is challenging.
method Train rank-one perturbations and add them to weight matrices infrequently, keeping track of inverses and determinants.
result Invertible linear layers improve mixing and mode separation in normalizing flows.
SPTN uses invertible transformations to improve sum-product networks.
problem Improving inference efficiency and tractability in sum-product networks.
method Integrates invertible transformations into sum-product networks (SPNs).
result SPTNs with Gaussian leaves and affine transformations are as tractable as SPNs.
INNs can approximate diverse functions despite layer restrictions.
problem Can INNs approximate sufficiently diverse functions?
method Developed a theoretical framework based on differential geometry to simplify the approximation problem of diffeomorphisms.
result INNs have the universal approximation property.
Generative Adversarial Networks (GANs) play an increasingly important role in machine learning. However, there is one fundamental issue hindering their practical applications: the absence of capability for encoding real-world samples. The conventional way of addressing this issue is to learn an encoder for GAN via Vari…
In this manuscript, we investigate deep invertible networks for EEG-based brain signal decoding and find them to generate realistic EEG signals as well as classify novel signals above chance. Further ideas for their regularization towards better decoding accuracies are discussed.
Improves conditional distribution modeling with simpler training.
problem Analyzing inverse problems with invertible neural networks.
method Uses normalizing flows to maximize posterior likelihood, incorporating conditioning.
result Easier training and natural framework for conditional generation.
JacNet learns Jacobians to enforce structure on derivatives for invertibility and Lipschitz functions.
problem Enforcing structure on derivatives of neural network mappings.
method Proposes using a neural network to directly learn the Jacobian of the input-output function, allowing control over derivative structure.
result Demonstrates learning invertible approximations to simple and 1-Lipschitz functions.
New invertible transformations improve flow-based generative models.
problem Improving flow-based generative models for better performance.
method Proposed new invertible transformations and coupling layers.
result New coupling layers achieve better results in IDF.
This paper proposes a new method to quantify uncertainty in reservoir characterization using invertible neural networks.
problem Quantifying uncertainty in reservoir characterization models.
method Training an invertible neural network to represent the posterior distribution of model parameters.
result The proposed method provides a more efficient and direct way to sample from the posterior distribution.
AIKAE enhances IKAE for long-term time series forecasting.
problem Limitation of dimension conservation in IKAE models.
method Augmented with a non-invertible encoder network.
result AIKAE improves long-term forecasting accuracy.
INVERT connects neural representations to human-understandable concepts.
problem Lack of understanding and statistical significance in existing explainability methods.
method Inverse Recognition (INVERT) approach that connects learned representations to human-understandable concepts.
result INVERT provides interpretable metrics and statistical significance for representation alignment.
Momentum ResNets improve ResNets' memory efficiency.
problem Memory inefficiency in deep residual neural networks (ResNets).
method Adding a momentum term to the forward rule of ResNets to make them invertible.
result Momentum ResNets can learn any linear mapping up to a multiplicative factor, improving memory efficiency.
Many recent invertible neural architectures are based on coupling block designs where variables are divided in two subsets which serve as inputs of an easily invertible (usually affine) triangular transformation. While such a transformation is invertible, its Jacobian is very sparse and thus may lack expressiveness. Th…
Neural ODEs and i-ResNet are recently proposed methods for enforcing invertibility of residual neural models. Having a generic technique for constructing invertible models can open new avenues for advances in learning systems, but so far the question of whether Neural ODEs and i-ResNets can model any continuous inverti…
SGNs use Hamiltonian mechanics for invertible deep generative modeling.
problem Efficient and exact likelihood evaluation for deep generative models.
method Symplectic structure in latent space, Hamiltonian dynamics for data generation.
result Exact likelihood evaluation without Jacobian calculations.
Monotonic neural networks have recently been proposed as a way to define invertible transformations. These transformations can be combined into powerful autoregressive flows that have been shown to be universal approximators of continuous probability distributions. Architectures that ensure monotonicity typically enfor…
Capsule networks improve performance on image classification tasks with fewer parameters.
problem Improving performance of capsule networks with fewer parameters.
method Inverted dot-product attention routing, Layer Normalization, concurrent iterative routing.
result Improves performance on benchmark datasets CIFAR-10 and CIFAR-100, and performs at-par with ResNet-18.
The paper explores how invertibility affects the complexity of encoder models in VAEs.
problem The complexity of the encoder model in VAEs when the generative map is invertible.
method Formalizes the concept of strong invertibility and analyzes the complexity of the encoder model.
result Strongly invertible generative maps allow for simpler encoder models, while non-invertible maps require exponentially larger encoders.
This research reverses feature visualization in neural networks to optimize for specific feature objectives.
problem The invertibility of feature visualization in neural networks is not well understood.
method The approach involves optimizing for the feature objective that generates the input used in feature visualization, using the gradient of a specific objective function.
result A closed-form solution is found to minimize the gradient, providing an alternative view on network sensitivity.
i-DenseNets improve parameter efficiency and performance in density estimation.
problem Improving parameter efficiency and performance in density estimation models.
method Invertible Dense Networks (i-DenseNets) with learnable weighted concatenation and Concatenated LipSwish activation function.
result i-DenseNets outperform Residual Flows and other flow-based models in bits per dimension.
Several recent works have empirically observed that Convolutional Neural Nets (CNNs) are (approximately) invertible. To understand this approximate invertibility phenomenon and how to leverage it more effectively, we focus on a theoretical explanation and develop a mathematical model of sparse signal recovery that is c…
Flowification enriches neural networks with an inverse pass and likelihood monitoring.
problem Neural networks lack an inverse pass and likelihood monitoring, limiting their generative capabilities.
method Introduce flowification, enriching neural networks with a stochastic inverse pass and likelihood monitoring.
result Certain neural network architectures can be enriched to fall under the generalized notion of a normalizing flow.
Incorporates matrix exponential into generative flows for improved performance.
problem Improving generative flow models for better density estimation.
method Integrates matrix exponential into generative flows, proposing new layers and modifying network architecture.
result The proposed model achieves great performance on density estimation.
It is widely believed that the success of deep convolutional networks is based on progressively discarding uninformative variability about the input with respect to the problem at hand. This is supported empirically by the difficulty of recovering images from their hidden representations, in most commonly used network …
Survival MDN uses invertible functions to speed up survival analysis models.
problem Training neural ODEs for survival analysis is computationally expensive.
method Survival MDN applies an invertible positive function to MDN outputs.
result Survival MDN outperforms or matches other models on concordance, Brier score, and log-likelihood.
A new method combines generative models and regularization to prevent forgetting in continual learning.
problem Catastrophic forgetting in neural networks when learning new tasks.
method Uses a normalizing flow as a generative model to keep past data embeddings and regularize them.
result Performs favorably compared to existing methods with constant memory overhead.
GONs improve predictions of maximizers from noisy black-box functions.
problem Estimating maximizers of noisy black-box functions.
method Global Optimization Networks (GONs) composed of invertible and unimodal functions.
result GONs outperform convex fits, GPR, and DNNs in prediction accuracy.
Hydrogen atom confined in an inverted-Gaussian potential, with detailed numerical methods and results.
problem Studying hydrogen atom in a specific potential.
method Three numerical methods: Lagrange-mesh, fourth order finite differences, and finite element method.
result Accurate numerical results for hydrogen atom energies and eigenfunctions, improving previous literature.
IZF uses flow-based models to overcome ZSL limitations.
problem Hardness of training ZSL models and limited generation quality.
method IZF incorporates flow-based models to learn factorized data embeddings and generates samples.
result IZF significantly outperforms existing methods on ZSL benchmarks.
Adversarial training enhances model transferability without sacrificing accuracy.
problem The principle of minimal information in classification models is challenged by adversarial training.
method Investigation of the dual relationship between adversarial training and information theory.
result Adversarial training improves linear transferability and introduces a trade-off between transferability and source task accuracy.
InvGAN solves GAN inference problem without training data.
problem Inferring latent variables in GANs without access to training data.
method Training an encoder network to invert a pre-trained generator.
result Approximate inversion of latent codes from GAN generations.
In this work we compute lower Lipschitz bounds of ℓp pooling operators for p=1,2,∞ as well as ℓp pooling operators preceded by half-rectification layers. These give sufficient conditions for the design of invertible neural network layers. Numerical experiments on MNIST and image patches confirm tha…