A new method directly encodes data into latent space using gradient flow.
problem Suboptimal representations in physical sciences due to encoder inversion.
method Decoder-only approach using gradient flow and ODEs, avoiding integrals.
result Superior data efficiency and explicit encoding compared to traditional autoencoders.
This paper considers the problem of implementing large-scale gradient descent algorithms in a distributed computing setting in the presence of {\em straggling} processors. To mitigate the effect of the stragglers, it has been previously proposed to encode the data with an erasure-correcting code and decode at the maste…
Gradient flow autoencoder improves data efficiency over traditional autoencoders.
problem Sub-optimal latent space representations in autoencoders.
method Gradient flow through ODE with adaptive step size for optimization.
result Gradient flow autoencoder achieves higher data efficiency.
Enhances GBDT robustness with one-hot encoding and regularization.
problem Low robustness of GBDT models against covariate perturbation.
method One-hot encoding to linear framework, risk decomposition, L1 or L2 regularization. result Regularization enhances GBDT robustness.
Unbiased gradient estimation improves VAE performance.
problem Training VAEs via maximum likelihood is difficult due to intractable integrals.
method Introduced unbiased estimators of the log-likelihood gradient using coupled Markov chains.
result Unbiased estimators lead to better predictive performance in VAEs.
Conformer encoder reverses sequence in time dimension, affecting decoder training.
problem Reversal of sequence in Conformer encoder impacts decoder training.
method Analyzed initial behavior of decoder cross-attention and proposed methods to avoid flipping.
result Self-attention module of Conformer starts dominating, allowing only reversed information to pass.
New method for scalable set encoding with unbiased gradient approximation.
problem Limited expressive power and large set training issues in set functions.
method Universally MBC (UMBC) class of set functions and efficient MBC training algorithm.
result Unbiased approximation of full set gradient with constant memory overhead.
This paper improves deep learning models' accuracy with differential privacy using gradient encoding and denoising.
problem Deep learning models leak sensitive information about their training datasets.
method Gradient encoding to map gradients to a smaller vector space, and denoising for post-processing.
result Our technique achieves better model accuracy with differential privacy guarantees compared to state-of-the-art methods.
Paper analyzes dataset distillation for efficient encoding of task-relevant information.
problem Efficiently encoding task-relevant information from gradient-based learning of non-linear tasks.
method Theoretical analysis of dataset distillation applied to two-layer neural networks with gradient-based training.
result Low-dimensional structure of the problem is efficiently encoded into distilled data, reproducing a model with high generalization ability.
Improves shared encoder representations for better multi-task learning performance.
problem Improving quality of shared encoder representations in multi-task learning.
method Dummy Gradient norm Regularization (DGR) to decrease gradient norm of dummy task-specific predictors.
result DGR improves multi-task prediction performances and superior performance compared to existing methods.
Paper formalizes and analyzes a new bound for variational inference.
problem Lack of theoretical guarantees in variational algorithms.
method Introduces VR-IWAE bound, a generalization of IWAE.
result VR-IWAE bound leads to unbiased gradient estimators.
Transformers learn causal structure through gradient descent on self-attention mechanisms.
problem Understanding how transformers learn causal structure during training.
method In-context learning task and simplified two-layer transformer model.
result Gradient descent on a simplified transformer learns to encode latent causal graphs.
Novel PCA method for high-dimensional inverse problems.
problem Optimizing large-scale random fields with gradient information.
method Gradient-Sensitive Principal Component Analysis (Gradient-SPCA) that modifies PCA using objective function gradients.
result Improvements in encoding quality for objective function minimization and field distribution.
NES optimizes discrete structured VAEs effectively without gradient propagation.
problem Learning high-dimensional discrete latent spaces in generative models.
method Natural Evolution Strategies (NES) for gradient-free optimization of discrete structures.
result NES effectively optimizes discrete structured VAEs, comparable to gradient-based methods.
PE-SVI reduces SVI inference complexity by finding a suitable start point.
problem Complex posterior inference in graphical models leads to suboptimal learning.
method PE-SVI uses a pseudo-encoded start point to reduce gradient steps and step sizes.
result PE-SVI achieves the same ELBo objective as SVI with less than 1% of the required steps.
MetaFun learns functional representations for meta-learning.
problem Few-shot classification on large-scale datasets.
method Functional encoder-decoder approach with iterative updates.
result State-of-the-art performance on miniImageNet and tieredImageNet.
We prove that the evolution of weight vectors in online gradient descent can encode arbitrary polynomial-space computations, even in very simple learning settings. Our results imply that, under weak complexity-theoretic assumptions, it is impossible to reason efficiently about the fine-grained behavior of online gradie…
Federated learning is a distributed learning method to train a shared model by aggregating the locally-computed gradient updates. In federated learning, bandwidth and privacy are two main concerns of gradient updates transmission. This paper proposes an end-to-end encrypted neural network for gradient updates transmiss…
New hashing method improves document retrieval precision.
problem Efficiently retrieving similar documents from large text databases.
method Pairwise supervised hashing with Bernoulli VAE and unbiased gradient estimator.
result Superior performance compared to existing methods.
This paper will explore the use of autoencoders for semantic hashing in the context of Information Retrieval. This paper will summarize how to efficiently train an autoencoder in order to create meaningful and low-dimensional encodings of data. This paper will demonstrate how computing and storing the closest encodings…
Image compression is an essential approach for decreasing the size in bytes of the image without deteriorating the quality of it. Typically, classic algorithms are used but recently deep-learning has been successfully applied. In this work, is presented a deep super-resolution work-flow for image compression that maps …
This research compares two encoding methods for categorical attributes in machine learning, affecting model fairness.
problem The impact of encoding protected categorical attributes on fairness in machine learning models.
method Comparison of one-hot encoding and target encoding methods.
result Target encoding can lead to more unfair models compared to one-hot encoding due to induced bias.
New bounds enable training of probabilistic models for deep networks.
problem Training scalable latent variable models for deep networks.
method Introducing new variational bounds for specific output layers of neural networks.
result Analytical bounds for certain output layers allow training without re-parameterization or Monte Carlo approximations.
Regularized target encoding beats traditional methods for high cardinality features in ML.
problem Efficiently encoding high cardinality categorical variables for ML algorithms.
method Regularized target encoding compared to traditional encodings like integer and one-hot encoding.
result Regularized target encoding consistently provided the best results in a large-scale benchmark experiment.
Performance of distributed optimization and learning systems is bottlenecked by "straggler" nodes and slow communication links, which significantly delay computation. We propose a distributed optimization framework where the dataset is "encoded" to have an over-complete representation with built-in redundancy, and the …
We extend contrastive learning theory for multiway classification and prove convergence guarantees.
problem Efficient self-supervised training for multiway classification tasks.
method Contrastive representation learning with multiple negative samples and convergence guarantees for gradient descent.
result Convergence guarantees for contrastive learning with gradient descent of an overparametrized encoder.
A new method for target propagation using iterative approximations converges fast and is more biologically plausible.
problem Improving target propagation methods for neural networks.
method Iterative approximate inverses and local auto-encoders.
result The method converges exponentially fast under certain conditions.
We provide theoretical and empirical evidence that using tighter evidence lower bounds (ELBOs) can be detrimental to the process of learning an inference network by reducing the signal-to-noise ratio of the gradient estimator. Our results call into question common implicit assumptions that tighter ELBOs are better vari…
This paper proposes an unsupervised learning method to solve heat equations on chips.
problem Critical need for solving heat transfer equations on chips for 5G and AI.
method Hybrid framework of Auto Encoder and Image Gradient for unsupervised learning.
result Framework can solve heat transfer problems with a single training process and predict unseen cases.
MIM learns joint distributions with mutual information and low divergence.
problem Learning joint distributions over observations and latent variables.
method Probabilistic auto-encoder with three design principles: low divergence, high mutual information, and low marginal entropy.
result MIM learns representations with high mutual information, consistent encoding and decoding distributions, effective latent clustering, and comparable data log likelihood to VAE.
Variational Auto-Encoders (VAEs) have become very popular techniques to perform inference and learning in latent variable models as they allow us to leverage the rich representational power of neural networks to obtain flexible approximations of the posterior of latent variables as well as tight evidence lower bounds (…
Gradient boosting method enforced with individual fairness.
problem Enforcing fairness in machine learning models.
method Functional gradient descent on robust loss function.
result Algorithm converges globally and generalizes.
Affine spiking neural networks learn efficiently and generalize well.
problem Learning with spiking neural networks, especially with positive weights.
method Affine encoders and decoders, continuous parameter dependence, gradient-based training.
result Affine spiking neural networks can approximate shallow ReLU networks and generalize well.
Gradient boosted models are a fundamental machine learning technique. Robustness to small perturbations of the input is an important quality measure for machine learning models, but the literature lacks a method to prove the robustness of gradient boosted models. This work introduces VeriGB, a tool for quantifying the …
Researchers interpret SGD using diffusion metrics for clearer geometric understanding.
problem Elusiveness of geometrical significance in stochastic gradient descent.
method Study a deterministic model with geodesics of diffusion metrics.
result Establishes parallel with General Relativity models.
Quantum networks offer exponential communication savings for large machine learning models.
problem Training and inference of large models require efficient communication.
method Quantum encoding and gradient descent for distributed computation.
result Exponential reduction in communication for gradient descent on quantum networks.
We present the perceptor gradients algorithm -- a novel approach to learning symbolic representations based on the idea of decomposing an agent's policy into i) a perceptor network extracting symbols from raw observation data and ii) a task encoding program which maps the input symbols to output actions. We show that t…
A graph VAE framework optimizes neural architectures in a continuous space.
problem Discovering efficient neural architectures in a discrete space.
method Graph VAE framework with VAE and GNN components, joint learning of predictors and decoders.
result The framework discovers powerful neural architectures with both excellent performance and high computational efficiency.
Inference models are a key component in scaling variational inference to deep latent variable models, most notably as encoder networks in variational auto-encoders (VAEs). By replacing conventional optimization-based inference with a learned model, inference is amortized over data examples and therefore more computatio…
To address the challenge of backpropagating the gradient through categorical variables, we propose the augment-REINFORCE-swap-merge (ARSM) gradient estimator that is unbiased and has low variance. ARSM first uses variable augmentation, REINFORCE, and Rao-Blackwellization to re-express the gradient as an expectation und…
Gradient descent on LSE objectives implicitly performs EM, leading to collapse without volume control.
problem Gradient collapse in autoencoders without volume control.
method Introduced a single-layer encoder with an LSE objective and InfoMax regularization for volume control.
result Gradient--responsibility identity holds exactly; LSE alone collapses; variance prevents dead components; decorrelation prevents redundancy.
In this note we present a generative model of natural images consisting of a deep hierarchy of layers of latent random variables, each of which follows a new type of distribution that we call rectified Gaussian. These rectified Gaussian units allow spike-and-slab type sparsity, while retaining the differentiability nec…
We study the dynamics of the vector field on an open surface given by the gradient of a Green's function. This dynamical approach enables us to show that this field induces an invariant decomposition of the surface as the union of a disk and a 1-skeleton that encodes the topology of the surface. We analyze the structur…
Conventional embedding methods directly associate each symbol with a continuous embedding vector, which is equivalent to applying a linear transformation based on a "one-hot" encoding of the discrete symbols. Despite its simplicity, such approach yields the number of parameters that grows linearly with the vocabulary s…
StructureBoost improves gradient boosting for complex categorical variables efficiently.
problem Efficiently handling complex categorical variables with known structure.
method Two methods to overcome computational obstacles in SCDT enumeration for structured categorical variables.
result StructureBoost outperforms existing packages on complex categorical problems.
New estimator reduces variance in discrete random variables.
problem Estimating gradients for discrete random variables with reduced variance.
method Sampling without replacement and Rao-Blackwellization.
result Our estimator is the most consistent gradient estimator across different entropy settings.
New methods improve Fisher Matrix approximations for neural networks at low cost.
problem High cost of solving Fisher Information Matrix (FIM) in neural networks.
method Direct minimization via Kronecker product singular value decomposition.
result Improved approximations to FIM provide more accurate and faster optimization.
A new method learns graph compression from data.
problem Graphs lack ordering, making conventional compression algorithms ineffective.
method Partition and Code framework: decompose, learn, encode.
result PnC achieves compression gains that grow with graph size.