Picture for Borjan Geshkovski

Borjan Geshkovski

Measure-to-measure interpolation using Transformers

Add code
Nov 07, 2024
Viaarxiv icon

Dynamic metastability in the self-attention model

Add code
Oct 09, 2024
Figure 1 for Dynamic metastability in the self-attention model
Figure 2 for Dynamic metastability in the self-attention model
Figure 3 for Dynamic metastability in the self-attention model
Figure 4 for Dynamic metastability in the self-attention model
Viaarxiv icon

A mathematical perspective on Transformers

Add code
Dec 22, 2023
Figure 1 for A mathematical perspective on Transformers
Figure 2 for A mathematical perspective on Transformers
Figure 3 for A mathematical perspective on Transformers
Figure 4 for A mathematical perspective on Transformers
Viaarxiv icon

The emergence of clusters in self-attention dynamics

Add code
May 17, 2023
Viaarxiv icon

Turnpike in optimal control of PDEs, ResNets, and beyond

Add code
Feb 08, 2022
Viaarxiv icon

Sparse approximation in learning via neural ODEs

Add code
Feb 26, 2021
Figure 1 for Sparse approximation in learning via neural ODEs
Figure 2 for Sparse approximation in learning via neural ODEs
Figure 3 for Sparse approximation in learning via neural ODEs
Figure 4 for Sparse approximation in learning via neural ODEs
Viaarxiv icon

Large-time asymptotics in deep learning

Add code
Aug 06, 2020
Figure 1 for Large-time asymptotics in deep learning
Figure 2 for Large-time asymptotics in deep learning
Figure 3 for Large-time asymptotics in deep learning
Figure 4 for Large-time asymptotics in deep learning
Viaarxiv icon