Chulhee Yun

is this you? claim profile

0 followers
MIT

Provable Memorization via Deep Neural Networks using Sub-linear Parameters

It is known that Θ(N) parameters are sufficient for neural networks to m...
10/26/2020 ∙ by Sejun Park, et al. ∙ 0 ∙ share

read it
A Unifying View on Implicit Bias in Training Linear Neural Networks

We study the implicit bias of gradient flow (i.e., gradient descent with...
10/06/2020 ∙ by Chulhee Yun, et al. ∙ 5 ∙ share

read it
Minimum Width for Universal Approximation

The universal approximation property of width-bounded networks has been ...
06/16/2020 ∙ by Sejun Park, et al. ∙ 0 ∙ share

read it
SGD with shuffling: optimal rates without component convexity and large epoch requirements

We study without-replacement SGD for solving finite-sum optimization pro...
06/12/2020 ∙ by Kwangjun Ahn, et al. ∙ 0 ∙ share

read it
O(n) Connections are Expressive Enough: Universal Approximability of Sparse Transformers

Transformer networks use pairwise attention to compute contextual embedd...
06/08/2020 ∙ by Chulhee Yun, et al. ∙ 5 ∙ share

read it
Low-Rank Bottleneck in Multi-head Attention Models

Attention based Transformer architecture has enabled significant advance...
02/17/2020 ∙ by Srinadh Bhojanapalli, et al. ∙ 9 ∙ share

read it
Are Transformers universal approximators of sequence-to-sequence functions?

Despite the widespread adoption of Transformer models for NLP tasks, the...
12/20/2019 ∙ by Chulhee Yun, et al. ∙ 24 ∙ share

read it
Are deep ResNets provably better than linear predictors?

Recently, a residual network (ResNet) with a single residual block has b...
07/09/2019 ∙ by Chulhee Yun, et al. ∙ 2 ∙ share

read it
Finite sample expressive power of small-width ReLU networks

We study universal finite sample expressivity of neural networks, define...
10/17/2018 ∙ by Chulhee Yun, et al. ∙ 12 ∙ share

read it
Efficiently testing local optimality and escaping saddles for ReLU networks

We provide a theoretical algorithm for checking local optimality and esc...
09/28/2018 ∙ by Chulhee Yun, et al. ∙ 18 ∙ share

read it
A Critical View of Global Optimality in Deep Learning

We investigate the loss surface of deep linear and nonlinear neural netw...
02/10/2018 ∙ by Chulhee Yun, et al. ∙ 0 ∙ share

read it
Global optimality conditions for deep neural networks

We study the error landscape of deep linear and nonlinear neural network...
07/08/2017 ∙ by Chulhee Yun, et al. ∙ 0 ∙ share

read it

Chulhee Yun

Featured Co-authors

Suvrit Sra

Jinwoo Shin

Sanjiv Kumar

Ali Jadbabaie

Sashank J Reddi

Srinadh Bhojanapalli

Ankit Singh Rawat

Hossein Mobahi

Sejun Park

Kwangjun Ahn

Jaeho Lee

Provable Memorization via Deep Neural Networks using Sub-linear Parameters

A Unifying View on Implicit Bias in Training Linear Neural Networks

Minimum Width for Universal Approximation

SGD with shuffling: optimal rates without component convexity and large epoch requirements

O(n) Connections are Expressive Enough: Universal Approximability of Sparse Transformers

Low-Rank Bottleneck in Multi-head Attention Models

Are Transformers universal approximators of sequence-to-sequence functions?

Are deep ResNets provably better than linear predictors?

Finite sample expressive power of small-width ReLU networks

Efficiently testing local optimality and escaping saddles for ReLU networks

A Critical View of Global Optimality in Deep Learning

Global optimality conditions for deep neural networks