Dmitry Vetrov

Часть публикаций из профиля, по году от новых к старым. Это не полный список последних работ автора; работы без года — в конце.

• Can Stationary Distributions of Scale-Invariant Neural Networks Be Described by the Thermodynamics of an Ideal Gas?2025
• Adaptive Destruction Processes for Diffusion Samplers2025
• Smoothie: Smoothing Diffusion on Token Embeddings for Text Generation2025
• TEncDM: Understanding the Properties of the Diffusion Model in the Space of Language Model Encodings2025
• Streaming Generation of Co-Speech Gestures via Accelerated Rolling Diffusion2025
• Averaging Weights Leads to Wider Optima and Better Generalization2018 · аннотациясвернуть

  Deep neural networks are typically trained by optimizing a loss function with an SGD variant, in conjunction with a decaying learning rate, until convergence. We show that simple averaging of multiple points along the…

• Spatially Adaptive Computation Time for Residual Networks2017 · аннотациясвернуть

  This paper proposes a deep learning architecture based on Residual Network that dynamically adjusts the number of executed layers for the regions of the image. This architecture is end-to-end trainable, deterministic…

• Variational Dropout Sparsifies Deep Neural Networks2017 · аннотациясвернуть

  We explore a recently proposed Variational Dropout technique that provided an elegant Bayesian interpretation to Gaussian Dropout. We extend Variational Dropout to the case when dropout rates are unbounded, propose a…