Markov chain Phase-Type decoders in VAEs overcome the structural inability of Gaussian-Lipschitz models to produce heavy-tailed outputs, cutting tail KS distance by up to 6x and extreme quantile error by up to 10x on synthetic Pareto data.
Cont , title =
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
representative citing papers
Polynomial-time SDP and ellipsoid-based approximation of Kolmogorov widths yields efficient robust detection boundaries matching upper bounds up to polylog factors for structured constrained signals.
In competitive threshold models, fixed thresholds lead to fragmentation at high resistance while fluid thresholds produce winner-takes-all consensus, with discontinuous transitions in both cases.
Rough-path market models satisfying no-controlled-free-lunch reduce admissible drivers to Itô lifts of Brownian motion (up to time change) once signature-type strategies are allowed.
SGD weight matrices are portfolio allocations whose spectra transition from Marchenko-Pastur to inverse-Wishart, unified by a Spectral Invariance Theorem that preserves singular-value distributions under isotropic perturbations.
Wealth tax neutrality holds under stochastic volatility and Epstein-Zin utility but fails for HARA preferences, while real taxes introduce distortions through non-uniform assessment, price effects, progressive thresholds that boost risk-taking near boundaries, and labor supply responses.
citing papers explorer
-
Markov Chain Decoders Overcome the Heavy-Tail Limitations of Lipschitz Generative Models
Markov chain Phase-Type decoders in VAEs overcome the structural inability of Gaussian-Lipschitz models to produce heavy-tailed outputs, cutting tail KS distance by up to 6x and extreme quantile error by up to 10x on synthetic Pareto data.
-
Efficient Robust Constrained Signal Detection via Kolmogorov Width Approximations
Polynomial-time SDP and ellipsoid-based approximation of Kolmogorov widths yields efficient robust detection boundaries matching upper bounds up to polylog factors for structured constrained signals.
-
Competitive Social Mobilization in Threshold Models of Collective Action
In competitive threshold models, fixed thresholds lead to fragmentation at high resistance while fluid thresholds produce winner-takes-all consensus, with discontinuous transitions in both cases.
-
Unbiased Rough Integrators and No Free Lunch in Rough-Path-Based Market Models
Rough-path market models satisfying no-controlled-free-lunch reduce admissible drivers to Itô lifts of Brownian motion (up to time change) once signature-type strategies are allowed.
-
Spectral Portfolio Theory: From SGD Weight Matrices to Wealth Dynamics
SGD weight matrices are portfolio allocations whose spectra transition from Marchenko-Pastur to inverse-Wishart, unified by a Spectral Invariance Theorem that preserves singular-value distributions under isotropic perturbations.
-
Extensions to the Wealth Tax Neutrality Framework
Wealth tax neutrality holds under stochastic volatility and Epstein-Zin utility but fails for HARA preferences, while real taxes introduce distortions through non-uniform assessment, price effects, progressive thresholds that boost risk-taking near boundaries, and labor supply responses.