Alert button
Picture for Andrej Risteski

Andrej Risteski

Alert button

Statistical Efficiency of Score Matching: The View from Isoperimetry

Oct 03, 2022
Frederic Koehler, Alexander Heckett, Andrej Risteski

Figure 1 for Statistical Efficiency of Score Matching: The View from Isoperimetry
Figure 2 for Statistical Efficiency of Score Matching: The View from Isoperimetry
Figure 3 for Statistical Efficiency of Score Matching: The View from Isoperimetry
Figure 4 for Statistical Efficiency of Score Matching: The View from Isoperimetry
Viaarxiv icon

Pitfalls of Gaussians as a noise distribution in NCE

Oct 01, 2022
Holden Lee, Chirag Pabbaraju, Anish Sevekari, Andrej Risteski

Figure 1 for Pitfalls of Gaussians as a noise distribution in NCE
Viaarxiv icon

Contrasting the landscape of contrastive and non-contrastive learning

Mar 29, 2022
Ashwini Pokle, Jinjin Tian, Yuchen Li, Andrej Risteski

Figure 1 for Contrasting the landscape of contrastive and non-contrastive learning
Figure 2 for Contrasting the landscape of contrastive and non-contrastive learning
Figure 3 for Contrasting the landscape of contrastive and non-contrastive learning
Figure 4 for Contrasting the landscape of contrastive and non-contrastive learning
Viaarxiv icon

Continual learning: a feature extraction formalization, an efficient algorithm, and fundamental obstructions

Mar 27, 2022
Binghui Peng, Andrej Risteski

Viaarxiv icon

Masked prediction tasks: a parameter identifiability view

Feb 18, 2022
Bingbin Liu, Daniel Hsu, Pradeep Ravikumar, Andrej Risteski

Viaarxiv icon

Sampling Approximately Low-Rank Ising Models: MCMC meets Variational Methods

Feb 17, 2022
Frederic Koehler, Holden Lee, Andrej Risteski

Viaarxiv icon

Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization

Feb 14, 2022
Elan Rosenfeld, Pradeep Ravikumar, Andrej Risteski

Figure 1 for Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization
Figure 2 for Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization
Figure 3 for Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization
Figure 4 for Domain-Adjusted Regression or: ERM May Already Learn Features Sufficient for Out-of-Distribution Generalization
Viaarxiv icon

Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias

Dec 13, 2021
Frederic Koehler, Viraj Mehta, Andrej Risteski, Chenghui Zhou

Figure 1 for Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias
Figure 2 for Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias
Figure 3 for Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias
Figure 4 for Variational autoencoders in the presence of low-dimensional data: landscape and implicit bias
Viaarxiv icon

Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation

Oct 21, 2021
Bingbin Liu, Elan Rosenfeld, Pradeep Ravikumar, Andrej Risteski

Figure 1 for Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation
Figure 2 for Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation
Figure 3 for Analyzing and Improving the Optimization Landscape of Noise-Contrastive Estimation
Viaarxiv icon

The Effects of Invertibility on the Representational Complexity of Encoders in Variational Autoencoders

Jul 09, 2021
Divyansh Pareek, Andrej Risteski

Viaarxiv icon