Picture for Jacob M. Springer

Jacob M. Springer

A Little Robustness Goes a Long Way: Leveraging Universal Features for Targeted Transfer Attacks

Jun 03, 2021
Jacob M. Springer, Melanie Mitchell, Garrett T. Kenyon

Figure 1 for A Little Robustness Goes a Long Way: Leveraging Universal Features for Targeted Transfer Attacks

Figure 2 for A Little Robustness Goes a Long Way: Leveraging Universal Features for Targeted Transfer Attacks

Figure 3 for A Little Robustness Goes a Long Way: Leveraging Universal Features for Targeted Transfer Attacks

Figure 4 for A Little Robustness Goes a Long Way: Leveraging Universal Features for Targeted Transfer Attacks

* 25 pages, 13 figures, 3 tables

Access Paper or Ask Questions

Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers

Feb 09, 2021
Jacob M. Springer, Melanie Mitchell, Garrett T. Kenyon

Figure 1 for Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers

Figure 2 for Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers

Figure 3 for Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers

Figure 4 for Adversarial Perturbations Are Not So Weird: Entanglement of Robust and Non-Robust Features in Neural Network Classifiers

* 20 pages, 14 figures, 6 tables

Access Paper or Ask Questions

STRATA: Building Robustness with a Simple Method for Generating Black-box Adversarial Attacks for Models of Code

Sep 28, 2020
Jacob M. Springer, Bryn Marie Reinstadler, Una-May O'Reilly

Figure 1 for STRATA: Building Robustness with a Simple Method for Generating Black-box Adversarial Attacks for Models of Code

Figure 2 for STRATA: Building Robustness with a Simple Method for Generating Black-box Adversarial Attacks for Models of Code

Figure 3 for STRATA: Building Robustness with a Simple Method for Generating Black-box Adversarial Attacks for Models of Code

Figure 4 for STRATA: Building Robustness with a Simple Method for Generating Black-box Adversarial Attacks for Models of Code

* 13 pages, 3 figures, 10 tables

Access Paper or Ask Questions

It's Hard for Neural Networks To Learn the Game of Life

Sep 03, 2020
Jacob M. Springer, Garrett T. Kenyon

Figure 1 for It's Hard for Neural Networks To Learn the Game of Life

Figure 2 for It's Hard for Neural Networks To Learn the Game of Life

Figure 3 for It's Hard for Neural Networks To Learn the Game of Life

Figure 4 for It's Hard for Neural Networks To Learn the Game of Life

* 12 pages, 6 figures

Access Paper or Ask Questions

Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples

Nov 20, 2018
Jacob M. Springer, Charles S. Strauss, Austin M. Thresher, Edward Kim, Garrett T. Kenyon

Figure 1 for Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples

Figure 2 for Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples

Figure 3 for Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples

Figure 4 for Classifiers Based on Deep Sparse Coding Architectures are Robust to Deep Learning Transferable Examples

* 8 pages, 8 figures, fixed typos

Access Paper or Ask Questions