[1]

Subspace of Maximum Curvature for Loss Landscape Stabilization

Slides
[2]

Adaptive Selection of the Gaussian Sampling Scale for Monte Carlo Estimation of Loss-Landscape Convergence

Slides
[3]

A mean-square criterion for loss-landscape convergence based on the leading eigenspace of the Hessian

SlidesVideo
[4]

Robust Convergence of Loss Landscapes through Distributional Averaging

SlidesVideo
[5]

Kandinsky: image and video diffusion models fine-tuning for personalization

SlidesVideo
[6]

Kandinsky: launching and fine-tuning image and video generation models

SlidesVideo
[7]

Sufficient sample size and its connection to the loss landscape convergence

Slides
[8]

Decoding visual stimuli from multimodal brain signals

Slides
[9]

Loss landscape convergence as an indicator of sufficient sample size

SlidesVideo
[10]

Hessian-based convergence analysis of convolutional network loss landscapes

Slides
[11]

Determining a sufficient sample size from the posterior distribution of model parameters

SlidesVideo
[12]

Spatial-temporal methods for time series analysis

SlidesVideo