... | @@ -5,7 +5,7 @@ |
... | @@ -5,7 +5,7 @@ |
|
* Figure 1 from [1]:
|
|
* Figure 1 from [1]:
|
|
![tcav_fig1](uploads/97e2298575d0a21290d89fd80c5f779a/tcav_fig1.png)
|
|
![tcav_fig1](uploads/97e2298575d0a21290d89fd80c5f779a/tcav_fig1.png)
|
|
* ⓐ user-defined set of examples for some concept $`C`$ (top-row, e.g. 'striped') + random examples (bottom row)
|
|
* ⓐ user-defined set of examples for some concept $`C`$ (top-row, e.g. 'striped') + random examples (bottom row)
|
|
* ⓑ labeled data examples for the studied class (e.g. zebras). Must correspond to a logit in the DNN. $k$ denotes the index of that logit.
|
|
* ⓑ labeled data examples for the studied class (e.g. zebras). Must correspond to a logit in the DNN. $`k`$ denotes the index of that logit.
|
|
* ⓒ DNN to be inspected. $`l`$ denotes the layer to hook into, i.e. the intermediate representation, and $`m`$ is the flattened intermediate representation.
|
|
* ⓒ DNN to be inspected. $`l`$ denotes the layer to hook into, i.e. the intermediate representation, and $`m`$ is the flattened intermediate representation.
|
|
* ⓓ Linear SVM classifier, hyperplane seperates in-concept and out-of-concept examples. The normal $`v_C^l`$ is the CAV, pointing towards the direction of the in-concept examples.
|
|
* ⓓ Linear SVM classifier, hyperplane seperates in-concept and out-of-concept examples. The normal $`v_C^l`$ is the CAV, pointing towards the direction of the in-concept examples.
|
|
* ⓔ Given an instance of the studied class (zebras), the conceptal sensitivity of the prediction for that instance towards a concept is quantified by the directional derivative $`S_{C,k,l}`$, that is, the gradient of the logit w.r.t. the intermediate representation in the direction of the concepts corresponding CAV (via dot product).
|
|
* ⓔ Given an instance of the studied class (zebras), the conceptal sensitivity of the prediction for that instance towards a concept is quantified by the directional derivative $`S_{C,k,l}`$, that is, the gradient of the logit w.r.t. the intermediate representation in the direction of the concepts corresponding CAV (via dot product).
|
... | | ... | |