Salient Imagenet

Class: 236 (Dobermann), Feature: 1313

This feature is core for class Dobermann. It is the most important feature for class Dobermann.

Below, we display the five Dobermann images that most highly activate the feature, along with the Neural Activation Map (heatmap) and Feature Attack for each image. These are the same as the images shown to crowdworkers to annotate this class-feature pair.

We additionally provide the full responses of the five crowdworkers that annotated this pair, as well as links to other related pairs (i.e. same class or same feature).

MTurk Worker Annotations

core

spurious

core

spurious

Main Focus	Confidence	Explanation
Main Object	5	Focus is on the face and legs of the doberman.
Main Object	4	The main obect is doberman whose some part got focus.
Main Object	5	The focus of the red region in the highlighted visual attributes is on doberman
Main Object	4	Focus is on body and face of doberman
Main Object	4	Focus is on the forelimbs of the doberman.

core

Related Class-Feature Pairs

Important Features for Dobermann

There are 4 features annotated as core and 1 features annotated as spurious for Dobermann. Core features include 220, 747, 1313, 1385. Spurious features include 1634.

Classes where Feature 1313 is among Most Important (i.e. top-5)

Feature 1313 is core for the following classes: Black And Tan Coonhound, Gordon Setter, Rottweiler, Dobermann, Miniature Pinscher, Greater Swiss Mountain Dog, Bernese Mountain Dog, Appenzeller Sennenhund, Entlebucher Sennenhund, Tibetan Mastiff.