Feature Visualization
By Deborah Silver. 1994
Subfields: Computer Vision and Pattern Recognition, Safety, Robustness & Interpretability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Ming-Hsuan Yang
- Dacheng Tao
- Alan L. Yuille
- Jitendra Malik
- Andrew Zisserman
- Trevor Darrell
- Andrea Vedaldi
In Safety, Robustness & Interpretability
- Aleksander Madry
- Been Kim
- Dimitris Tsipras
- Nicholas Carlini
- Matthias Hein
- Alexander Turner
- Logan Engstrom
- Andrew Ilyas
Citations in the corpus
- Adversarial Examples Are Not Bugs, They Are Features (2019, background): “It also suggests that approaches aiming to enhance the interpretability of a given model by enforcing “priors” for its explanation [MV15; OMS17; Smi+17] actually hide features that are “meaningful” and predictive to standard models.”
- Explanation Methods in Deep Learning: Users, Values, Concerns and Challenges (2018, background): “Fortunately, these problems have not escaped the attention of the ML/Deep Learning (DL) community (Zeng, 2016; Samek et al., 2017; Seifert et al., 2017; Olah et al., 2017; Hall et al., 2017; Montavon et al., 2018; Marcus, 2018; Doshi-Velez and Kim, 2017).”
- Interpreting Deep Classifier by Visual Distillation of Dark Knowledge (2018, background): “Another way to interpret deep neural networks is by means of visualization; for an overview, see Olah et al. (2017).”
- Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV) (2017, methodology): “Activation maximization techniques, such as Deep Dream or Lucid (Mordvintsev et al., 2015; Olah et al., 2017), are often used to visualize patterns that would maximally activate a neuron, set of neurons or random directions.”
- Visual interpretability for deep learning: a survey (2018, methodology): “[Olah et al., 2017] has provided a toolbox of existing techniques to visualize patterns encoded in different conv-layers of a pre-trained CNN.”
- Neural Prototype Trees for Interpretable Fine-grained Image Recognition (2020, background): “Others explain the internals of a model, with e.g . activation maximization [38, 40] to visualize neurons, deconvolution or upconvolution [11, 53] to explain layers, generating image exemplars [17] to explain the latent space, or concept activation vectors [25] to explain model sensitivity.”