A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
VALC provides concept-level interpretations of FLMs, overcoming word-level limitations.
This work interprets SFA through variational inference, relaxing linearity constraints.
DICCA maps multi-view data into a shared latent space with interpretable components.
We develop a framework for incorporating structured graphical models in the \emph{encoders} of variational autoencoders (VAEs) that allows us to induce interpretable representations through approximate variational inference. This allows us to both perform reasoning (e.g. classification) under the structural constraints…
IntroVAC learns interpretable latent subspaces for better image quality.
The paper investigates instance-based interpretation methods for VAEs.
ISVAE enhances interpretability in time series clustering using a novel filter bank.
We characterize how to vary the Abel-Jacobi map in terms of Schiffer variation. From this characterization, we will interpret the relation of hyperellipticity of curves with Schiffer variation and describe the deformation of elliptic solitons under Schiffer variation.
Neural Decomposition breaks down VAE latent structure for better interpretability.
ED-VAE improves VAEs by explicitly including entropy components in ELBO.
Lower bound for VAE training objective for binary data.
We propose a novel interpretation of the collapsed variational Bayes inference with a zero-order Taylor expansion approximation, called CVB0 inference, for latent Dirichlet allocation (LDA). We clarify the properties of the CVB0 inference by using the alpha-divergence. We show that the CVB0 inference is composed of two…
Interpretable machine learning has gained much attention recently. Briefness and comprehensiveness are necessary in order to provide a large amount of information concisely when explaining a black-box decision system. However, existing interpretable machine learning methods fail to consider briefness and comprehensiven…
SVGP KAN integrates sparse variational GP with KANs for scalable probabilistic inference.
We present a geometric interpretation of the integration-by-parts formula on an arbitrary vector bundle. As an application we give a new geometric formulation of higher-order variational calculus.
Geometric interpretation improves VAE performance and robustness.
In the paper "On Truncated Variation of Brownian Motion with Drift" (Bull. Pol. Acad. Sci. Math. 56 (2008), no.4, 267 - 281) we defined truncated variation of Brownian motion with drift, where is a standard Brownian motion. Truncated variation differs from regular variation by neglect…
SVGP KAN integrates uncertainty quantification into Kolmogorov-Arnold networks.
VICE embeds concepts in a vector space using human data.
In this paper, we provide an information-theoretic interpretation of the Vector Quantized-Variational Autoencoder (VQ-VAE). We show that the loss function of the original VQ-VAE can be derived from the variational deterministic information bottleneck (VDIB) principle. On the other hand, the VQ-VAE trained by the Expect…
This paper refines human labeling as a measurement process, revealing four sources of variation.
Extends V-IP framework to use LLMs for generating task-relevant concepts, improving interpretability and performance.
Physics models integrated into VAEs improve generative performance and extrapolation.
We show that unconverged stochastic gradient descent can be interpreted as a procedure that samples from a nonparametric variational approximate posterior distribution. This distribution is implicitly defined as the transformation of an initial distribution by a sequence of optimization updates. By tracking the change …
Method generates natural ECGs with 25 interpretable features.
Improved GPLVM model for single-cell RNA-seq data.
This paper presents a Semantic Attribute Modulation (SAM) for language modeling and style variation. The semantic attribute modulation includes various document attributes, such as titles, authors, and document categories. We consider two types of attributes, (title attributes and category attributes), and a flexible a…
pVAE combines physics and machine learning for robust data representations.
The paper studies stability of discrete planar curves using variational methods.
A new method for making interpretable predictions by sequentially asking questions, faster and more efficient.
In this paper we continue the investigation of Loday's Leibniz cohomology as a new invariant for differentiable manifolds. In particular the Leibniz coboundary of a k-tensor (in the sense of differential geometry) is computed in a local coordinate chart and then interpreted in terms of the calculus of variations. For e…
We construct a function of the edge-lengths of a triangulated surface whose variation under a rescaling of all the edges that meet at a vertex is the defect angle at that vertex. We interpret this function as a gravitational effective action on the triangulation, and the variation as a trace anomaly.
New method uses VAEs to generate financial correlation matrices for credit portfolio VaR analysis.
A 'Liouville structure' is a structure isomorphic to a cotangent vector fibration. A Liouville structure is an essential ingredient of every variational formulation of a physical theory. For reasons of interpretation the Liouville structure can not be replaced by the corresponding cotangent fibration. We give a precise…
In this paper, the dual Orlicz curvature measure is proposed and its basic properties are provided. A variational formula for the dual Orlicz-quermassintegral is established in order to give a geometric interpretation of the dual Orlicz curvature measure. Based on the established variational formula, a solution to the …
SAVAE uses deep learning for survival analysis, improving model performance and interpretability.
A framework for disentangling class-related and class-independent factors in data.
The standard interpretation of importance-weighted autoencoders is that they maximize a tighter lower bound on the marginal likelihood than the standard evidence lower bound. We give an alternate interpretation of this procedure: that it optimizes the standard variational lower bound, but using a more complex distribut…
Maintaining good cardiac function for as long as possible is a major concern for healthcare systems worldwide and there is much interest in learning more about the impact of different risk factors on cardiac health. The aim of this study is to analyze the impact of systolic blood pressure (SBP) on cardiac function whil…
Proposes a new model for directed graphs combining deep learning and latent variable models.
Electrocardiogram signals are omnipresent in medicine. A vital aspect in the analysis of this data is the identification and classification of heart beat types which is often done through automated algorithms. Advancements in neural networks and deep learning have led to a high classification accuracy. However, the fin…
VDA improves disentanglement of latent representations in complex signals.
Particle-based variational inference offers a flexible way of approximating complex posterior distributions with a set of particles. In this paper we introduce a new particle-based variational inference method based on the theory of semi-discrete optimal transport. Instead of minimizing the KL divergence between the po…
We develop a general variational inference method that preserves dependency among the latent variables. Our method uses copulas to augment the families of distributions used in mean-field and structured approximations. Copulas model the dependency that is not captured by the original variational distribution, and thus …
We introduce an elliptic regularization of the PDE system representing the isometric immersion of a surface in . The regularization is geometric, and has a natural variational interpretation.
Deep generative models are commonly used for generating images and text. Interpretability of these models is one important pursuit, other than the generation quality. Variational auto-encoder (VAE) with Gaussian distribution as prior has been successfully applied in text generation, but it is hard to interpret the mean…
UT module refines VAE latent space, improving disentanglement and interpretability.
Recent efforts on combining deep models with probabilistic graphical models are promising in providing flexible models that are also easy to interpret. We propose a variational message-passing algorithm for variational inference in such models. We make three contributions. First, we propose structured inference network…