Get our free extension to see links to code for papers anywhere online!

Add to Chrome

Add to Firefox

Picture for Yoonsuck Choe

Yoonsuck Choe

AdjointBackMapV2: Precise Reconstruction of Arbitrary CNN Unit's Activation via Adjoint Operators

Oct 04, 2021
Qing Wan, Yoonsuck Choe

Figure 1 for AdjointBackMapV2: Precise Reconstruction of Arbitrary CNN Unit's Activation via Adjoint Operators

Figure 2 for AdjointBackMapV2: Precise Reconstruction of Arbitrary CNN Unit's Activation via Adjoint Operators

Figure 3 for AdjointBackMapV2: Precise Reconstruction of Arbitrary CNN Unit's Activation via Adjoint Operators

Figure 4 for AdjointBackMapV2: Precise Reconstruction of Arbitrary CNN Unit's Activation via Adjoint Operators

* 19 pages, 8 figures

Access Paper or Ask Questions

Meaning Versus Information, Prediction Versus Memory, and Question Versus Answer

Jun 29, 2021
Yoonsuck Choe

* 14 pages

Access Paper or Ask Questions

AdjointBackMap: Reconstructing Effective Decision Hypersurfaces from CNN Layers Using Adjoint Operators

Dec 16, 2020
Qing Wan, Yoonsuck Choe

Figure 1 for AdjointBackMap: Reconstructing Effective Decision Hypersurfaces from CNN Layers Using Adjoint Operators

Figure 2 for AdjointBackMap: Reconstructing Effective Decision Hypersurfaces from CNN Layers Using Adjoint Operators

Figure 3 for AdjointBackMap: Reconstructing Effective Decision Hypersurfaces from CNN Layers Using Adjoint Operators

Figure 4 for AdjointBackMap: Reconstructing Effective Decision Hypersurfaces from CNN Layers Using Adjoint Operators

* Main paper + Appendix

Access Paper or Ask Questions

Emergence of Different Modes of Tool Use in a Reaching and Dragging Task

Dec 08, 2020
Khuong Nguyen, Yoonsuck Choe

Figure 1 for Emergence of Different Modes of Tool Use in a Reaching and Dragging Task

Figure 2 for Emergence of Different Modes of Tool Use in a Reaching and Dragging Task

Figure 3 for Emergence of Different Modes of Tool Use in a Reaching and Dragging Task

Figure 4 for Emergence of Different Modes of Tool Use in a Reaching and Dragging Task

* 11 pages, 10 figures, 14 pdf-embedded animations

Access Paper or Ask Questions

Action Recognition and State Change Prediction in a Recipe Understanding Task Using a Lightweight Neural Network Model

Jan 23, 2020
Qing Wan, Yoonsuck Choe

Figure 1 for Action Recognition and State Change Prediction in a Recipe Understanding Task Using a Lightweight Neural Network Model

* AAAI-2020 Student Abstract and Poster Program (Accept)

Access Paper or Ask Questions

Differentiable Pruning Method for Neural Networks

Apr 24, 2019
Jaedeok Kim, Chiyoun Park, Hyun-Joo Jung, Yoonsuck Choe

Figure 1 for Differentiable Pruning Method for Neural Networks

Figure 2 for Differentiable Pruning Method for Neural Networks

Figure 3 for Differentiable Pruning Method for Neural Networks

Figure 4 for Differentiable Pruning Method for Neural Networks

* Submitted to ICML Workshop

Access Paper or Ask Questions

How Compact?: Assessing Compactness of Representations through Layer-Wise Pruning

Jan 09, 2019
Hyun-Joo Jung, Jaedeok Kim, Yoonsuck Choe

Figure 1 for How Compact?: Assessing Compactness of Representations through Layer-Wise Pruning

Figure 2 for How Compact?: Assessing Compactness of Representations through Layer-Wise Pruning

Figure 3 for How Compact?: Assessing Compactness of Representations through Layer-Wise Pruning

Figure 4 for How Compact?: Assessing Compactness of Representations through Layer-Wise Pruning

* Accepted to AAAI 2019 Workshop on Network Interpretability for Deep Learning

Access Paper or Ask Questions

Comparing Sample-wise Learnability Across Deep Neural Network Models

Jan 08, 2019
Seung-Geon Lee, Jaedeok Kim, Hyun-Joo Jung, Yoonsuck Choe

Figure 1 for Comparing Sample-wise Learnability Across Deep Neural Network Models

Figure 2 for Comparing Sample-wise Learnability Across Deep Neural Network Models

Figure 3 for Comparing Sample-wise Learnability Across Deep Neural Network Models

* Accepted to AAAI 2019 Student Abstract

Access Paper or Ask Questions