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

Add to Chrome

Add to Firefox

Picture for Sara Hooker

Sara Hooker

When does loss-based prioritization fail?

Jul 16, 2021
Niel Teng Hu, Xinyu Hu, Rosanne Liu, Sara Hooker, Jason Yosinski

Figure 1 for When does loss-based prioritization fail?

Figure 2 for When does loss-based prioritization fail?

Figure 3 for When does loss-based prioritization fail?

Figure 4 for When does loss-based prioritization fail?

Access Paper or Ask Questions

Randomness In Neural Network Training: Characterizing The Impact of Tooling

Jun 22, 2021
Donglin Zhuang, Xingyao Zhang, Shuaiwen Leon Song, Sara Hooker

Figure 1 for Randomness In Neural Network Training: Characterizing The Impact of Tooling

Figure 2 for Randomness In Neural Network Training: Characterizing The Impact of Tooling

Figure 3 for Randomness In Neural Network Training: Characterizing The Impact of Tooling

Figure 4 for Randomness In Neural Network Training: Characterizing The Impact of Tooling

* 21 pages, 10 figures

Access Paper or Ask Questions

Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization

Feb 02, 2021
Kale-ab Tessera, Sara Hooker, Benjamin Rosman

Figure 1 for Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization

Figure 2 for Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization

Figure 3 for Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization

Figure 4 for Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization

Access Paper or Ask Questions

Characterising Bias in Compressed Models

Oct 06, 2020
Sara Hooker, Nyalleng Moorosi, Gregory Clark, Samy Bengio, Emily Denton

Figure 1 for Characterising Bias in Compressed Models

Figure 2 for Characterising Bias in Compressed Models

Figure 3 for Characterising Bias in Compressed Models

Figure 4 for Characterising Bias in Compressed Models

Access Paper or Ask Questions

The Hardware Lottery

Sep 21, 2020
Sara Hooker

Figure 1 for The Hardware Lottery

Figure 2 for The Hardware Lottery

Figure 3 for The Hardware Lottery

Figure 4 for The Hardware Lottery

Access Paper or Ask Questions

Estimating Example Difficulty using Variance of Gradients

Aug 26, 2020
Chirag Agarwal, Sara Hooker

Figure 1 for Estimating Example Difficulty using Variance of Gradients

Figure 2 for Estimating Example Difficulty using Variance of Gradients

Figure 3 for Estimating Example Difficulty using Variance of Gradients

Figure 4 for Estimating Example Difficulty using Variance of Gradients

* Accepted to Workshop on Human Interpretability in Machine Learning (WHI), ICML, 2020

Access Paper or Ask Questions

Selective Brain Damage: Measuring the Disparate Impact of Model Pruning

Nov 13, 2019
Sara Hooker, Aaron Courville, Yann Dauphin, Andrea Frome

Figure 1 for Selective Brain Damage: Measuring the Disparate Impact of Model Pruning

Figure 2 for Selective Brain Damage: Measuring the Disparate Impact of Model Pruning

Figure 3 for Selective Brain Damage: Measuring the Disparate Impact of Model Pruning

Figure 4 for Selective Brain Damage: Measuring the Disparate Impact of Model Pruning

Access Paper or Ask Questions

The State of Sparsity in Deep Neural Networks

Feb 25, 2019
Trevor Gale, Erich Elsen, Sara Hooker

Figure 1 for The State of Sparsity in Deep Neural Networks

Figure 2 for The State of Sparsity in Deep Neural Networks

Figure 3 for The State of Sparsity in Deep Neural Networks

Figure 4 for The State of Sparsity in Deep Neural Networks

Access Paper or Ask Questions

Evaluating Feature Importance Estimates

Jun 28, 2018
Sara Hooker, Dumitru Erhan, Pieter-Jan Kindermans, Been Kim

Figure 1 for Evaluating Feature Importance Estimates

Figure 2 for Evaluating Feature Importance Estimates

Figure 3 for Evaluating Feature Importance Estimates

Figure 4 for Evaluating Feature Importance Estimates

* presented at 2018 ICML Workshop on Human Interpretability in Machine Learning (WHI 2018), Stockholm, Sweden

Access Paper or Ask Questions