Boosts neural network performance by improving weight separability.
problem Improving the separability of weight vectors in neural networks.
method Proposes a new evaluation metric and feed-backward reconstruction loss to encourage weight separability.
result Improves visual recognition performance across various tasks.
Method separates data into class and style factors using semi-supervised learning.
problem Separating generative factors of data into class and style vectors.
method Independent Vector Variational Autoencoders with semi-supervised learning and independence term.
result Improves classification performance and generation controllability.
Study spectral properties of sparse random graphs to recover latent vectors.
problem Recovering latent vectors in sparse random geometric graphs.
method Analyzes spectral concentration and uses orthogonal polynomial expansions, decoupling, and matrix concentration.
result Sharpens spectral norm bounds and proves exact recovery for Gaussian mixture models.
We present a probabilistic language model for time-stamped text data which tracks the semantic evolution of individual words over time. The model represents words and contexts by latent trajectories in an embedding space. At each moment in time, the embedding vectors are inferred from a probabilistic version of word2ve…
A new model trains prior and encoder/decoder networks simultaneously for efficient generation.
problem Complex autoregressive prior in VQ-VAE models leads to slow generation.
method Builds a diffusion bridge between continuous and non-informative prior distributions.
result Model is competitive and efficient in optimization and sampling.
We develop necessary and sufficient conditions and a novel provably consistent and efficient algorithm for discovering topics (latent factors) from observations (documents) that are realized from a probabilistic mixture of shared latent factors that have certain properties. Our focus is on the class of topic models in …
Proposes a two-step method for sound source separation.
problem Improving sound source separation performance.
method First, learn a latent space transform. Second, train a separation module in the latent space.
result The proposed method achieves better performance than joint learning approaches.
Two VAE-based methods improve anomaly detection in semi-supervised settings.
problem Anomaly detection with limited labeled data.
method Intuitive idea of separating latent vectors for normal and outlier data, derived from probabilistic formulations.
result Marked improvement in outlier detection compared to state-of-the-art methods.
AR-Flow VAE improves blind source separation with flexible autoregressive priors.
problem Unsupervised blind source separation of latent signals from mixtures.
method AR-Flow VAE uses autoregressive flows to model latent sources, enhancing flexibility and capturing complex dependencies.
result AR-Flow VAE effectively separates latent sources, demonstrating improved performance over conventional methods.
Improves naturalness in TTS samples using quantized VAE and auto-regressive prosody.
problem Discontinuous and unnatural speech from standard VAE priors.
method Discretized latent features using vector quantization (VQ), and separately trained autoregressive (AR) prior model.
result Significantly improves naturalness in random sample generation.
Framework converts singer identity and vocal technique from non-parallel corpora.
problem Converts singer identity and vocal technique from non-parallel corpora.
method Uses variational autoencoders with separate encoders for singer identity and vocal technique.
result Successfully disentangles and converts singer identity and vocal technique.
The paper optimizes hyperplanes for binary classification in high-dimensional data with latent Gaussian mixtures.
problem Binary classification in high-dimensional data with latent Gaussian mixtures.
method Generalized least squares estimator for estimating the direction of the optimal separating hyperplane. Simple correction for intercept estimation.
result The procedure is minimax optimal in many scenarios and can retain the interpolation property.
The paper establishes a nearly-sharp statistical threshold for efficient learning in Latent MDPs with separated components.
problem Learning Latent Markov Decision Processes (LMDPs) with separated components.
method The paper considers various notions of separation and establishes a nearly-sharp statistical threshold for efficient learning. It also presents a quasi-polynomial algorithm with time complexity scaling in terms of the statistical threshold under a weaker assumption of separability under the optimal policy, and a near-matching time complexity lower bound under the exponential time hypothesis.
result Establishes a nearly-sharp statistical threshold for efficient learning in Latent MDPs with separated components.
Proposes a new condition to estimate latent variable causal graphs from observed data.
problem Estimating causal structures when observed variables are not the underlying causal variables.
method Introduces Generalized Independent Noise (GIN) condition and a recursive learning algorithm.
result Shows that GIN helps locate latent variables and identify their causal structure.
Paper shows how SFA fits into FBM framework for time series separation.
problem Identifying time series decomposition in flow-based models.
method Combining SFA and FBM to make time series decomposition identifiable.
result Time series decomposition becomes identifiable using SFA and FBM.
SAHMM-VAE separates sources adaptively using hidden Markov priors.
problem Unsupervised blind source separation.
method Source-wise adaptive Hidden Markov prior variational autoencoder.
result Different latent dimensions align with different source-specific temporal organizations.
Improved clustering and anomaly detection using latent space conditioning.
problem Anomaly detection on unlabeled data.
method Conditional latent space variational autoencoder (cLSVAE) that separates latent space by conditioning on data labels.
result The method outperforms typical variational autoencoders in clustering and anomaly detection.
Improves latent variable learning for complex data.
problem Expressive latent variables for model prediction on multi-component data.
method Dynamic Latent Separation method that distances data samples in the latent space.
result Enhances output diversity and provides interpretable representations.
Paper develops robust methods for panel data with latent groups, improving inference under group separation violations.
problem Inference in latent group panel models under group separation violations.
method Selective conditional inference approach to derive conditional distribution of coefficients given estimated group structure.
result Valid inference under violations of group separation, superior to traditional asymptotic methods.
A new text clustering method using NMF and LSA improves stability and performance.
problem Text data's large, sparse term-document matrix makes clustering difficult.
method Proposes a new feature agglomeration method based on NMF and deterministic K-Means initialization.
result Significantly improves clustering performance and stability.
StrEBM learns distinct latent components for better source separation.
problem Blind source separation with identifiable and decoupled latent components.
method Structured latent energy-based model with learnable structural biases.
result The model effectively recovers source components from mixed signals.
StrADiff separates sources from mixtures without labels, using structured priors.
problem Blind source separation of linear and nonlinear mixtures without labeled data.
method Structured Source-Wise Adaptive Diffusion Framework with Gaussian process priors.
result StrADiff can recover latent source trajectories in an unsupervised manner, especially stable in linear mixtures.
New graph types help identify complex relationships.
problem Understanding complex relationships in data.
method Introducing separable and essentially separable graphs to characterize and identify graphical models.
result Developed algorithms to identify equivalence classes of essentially separable graphs.
Latent representations are the essence of deep generative models and determine their usefulness and power. For latent representations to be useful as generative concept representations, their latent space must support latent space interpolation, attribute vectors and concept vectors, among other things. We investigate …
We assume that a high-dimensional datum, like an image, is a compositional expression of a set of properties, with a complicated non-linear relationship between the datum and its properties. This paper proposes a factorial mixture prior for capturing latent properties, thereby adding structured compositionality to deep…
Develops a new method to recover large latent tree models efficiently.
problem Inference of large latent tree structures from terminal node observations.
method Spectral Top-Down Recovery (STDR) using Fiedler vector partitioning.
result Proves statistical consistency and sample complexity for accurate tree recovery.
Unified framework for disentangled VAEs improves latent space interpretability.
problem Challenges in evaluating and interpreting latent representations, especially for diverse data types.
method Unified bfVAE framework, FVH-LT, DBSR-LS, GAS, LSSI.
result bfVAE provides more favorable trade-off between disentanglement and reconstruction.
UWM-JEPA predicts future scenarios in belief space, improving accuracy in partially observed environments.
problem Predicting future scenarios in partially observed environments with uncertainty.
method Introduces UWM-JEPA, a JEPA world model with a density-matrix latent and learned unitary predictor.
result UWM-JEPA achieves 0.77 accuracy on a hidden-velocity indicator task, outperforming LSTM-JEPA.
The paper uses deep neural networks to estimate economic models without separability restrictions.
problem Estimating economic models with complex interaction effects and non-separable restrictions.
method Uses deep neural networks as a nonparametric sieve to approximate regression functions from nonlinear latent variable models.
result Economic shape, sparsity, or separability restrictions are imposed more straightforwardly when a flexible latent variable model is used.
New method separates latent variables from noise in generative processes.
problem Reconstructing true latent variables from complex generative processes.
method General Incompressible-flow Network (GIN) for nonlinear ICA.
result Informative latent variables are automatically separated from noise.
Confidence-based filtering reveals latent structure in diffusion models.
problem Unclear latent structure in diffusion models.
method Confidence scores from a classifier.
result Class-relevant latent structure emerges under confidence-based filtering.
Paper distinguishes causal structures under latent confounding and selection bias.
problem Distinguishing causal relationships when latent variables and selection bias are present.
method Formulated selected-marginalized directed graphs (smDGs) to distinguish causal structures.
result Two causal structures are indistinguishable if they have the same selected-marginalized directed graph.
Quantizes latent space to improve disentanglement in models.
problem Learning disentangled representations from unlabeled data.
method Quantizes latent space into discrete code vectors with a learnable scalar codebook and applies high weight decay regularization.
result Quantized-latent autoencoder (QLAE) outperforms prior work in disentanglement without sacrificing data reconstruction.
We address the problem of causal discovery from data, making use of the recently proposed causal modeling framework of modular structural causal models (mSCM) to handle cycles, latent confounders and non-linearities. We introduce σ-connection graphs (σ-CG), a new class of mixed graphs (containing undirected, bidirected…
The support vector machine (SVM) is an important class of learning machines for function approach, pattern recognition, and time-serious prediction, etc. It maps samples into the feature space by so-called support vectors of selected samples, and then feature vectors are separated by maximum margin hyperplane. The pres…
The paper analyzes PLS-SVD in high-dimensional data integration, revealing its strengths and limitations.
problem Understanding the behavior of PLS-SVD in high-dimensional data integration.
method Analysis using random matrix theory and singular value decomposition.
result PLS-SVD exhibits counter-intuitive or limiting behavior in certain regimes and outperforms PCA when detecting common latent subspace.
This paper uses Factored Latent Analysis (FLA) to learn a factorized, segmental representation for observations of tracked objects over time. Factored Latent Analysis is latent class analysis in which the observation space is subdivided and each aspect of the original space is represented by a separate latent class mod…
Paper establishes identifiability conditions for a model with two latent vectors and auxiliary data.
problem Identifying conditions for a statistical model with two latent vectors and auxiliary data.
method Proposes a statistical model with two latent vectors and auxiliary data, establishing various identifiability conditions.
result Identifiability conditions reveal a dimensionality relation and link model indeterminacies to maximum link weights.
Novel method separates astrophysical components from noisy data.
problem Separating and reconstructing astrophysical components from noisy data.
method Latent-space field tension for automated component separation.
result High accuracy in reconstructing astrophysical components.
Polytopic Matrix Factorization models data as latent vectors from a polytope, maximizing determinant for identifiability.
problem Data decomposition with semi-structured latent vectors and polytope constraints.
method Model input data as latent vectors from a polytope, using determinant maximization for identifiability.
result Identifiability condition for polytopes with specific symmetry restrictions.
Develops a Bayesian non-parametric approach for signal separation with varying components.
problem Signal separation with varying components across different input locations.
method Augments Gaussian Process Latent Variable Models with weighted sums of pure component signals and incorporates priors for linear weights.
result Framework allows for non-linear variations in signals and incorporates useful priors for linear weights.
New approach separates VAE and GP for better molecular optimisation.
problem Optimizing complex structured domains like molecular spaces using VAEs.
method Decouples VAE for structure generation and GP for predictive modelling, combining them with a Bayesian update rule.
result Improves identification of high-potential candidates in molecular optimisation.
The paper predicts responses on out-of-sample nodes using latent positions on unknown curves.
problem Predicting responses on out-of-sample nodes with latent positions on unknown curves.
method Manifold learning and graph embedding technique using latent positions.
result Convergence guarantees for predicting responses on out-of-sample nodes.
We propose an unsupervised object matching method for relational data, which finds matchings between objects in different relational datasets without correspondence information. For example, the proposed method matches documents in different languages in multi-lingual document-word networks without dictionaries nor ali…
LTMs use latent vectors for efficient autoregressive generation.
problem Efficient autoregressive generation in language models.
method Dual-rate optimization in variational Bayes framework.
result LTMs achieve superior sample and parameter efficiency.
A new method for math reasoning that allows for iterative correction.
problem Standard reasoning models commit to each token and cannot recover from early errors.
method Generative framework with latent thought vectors for iterative self-correction.
result 30 rethinking iterations surpass baselines with 15 times more parameters.
The Generator of a Generative Adversarial Network (GAN) is trained to transform latent vectors drawn from a prior distribution into realistic looking photos. These latent vectors have been shown to encode information about the content of their corresponding images. Projecting input images onto the latent space of a GAN…
A new model designs molecules with desired properties.
problem Finding molecules with optimal chemical or biological properties.
method A latent prompt Transformer model with three components: latent vector, molecule generation, and property prediction.
result The model achieves state-of-the-art performance on molecule design tasks.