-
My question regards explanation methods for image classifiers.
Is there established way to compere two different "explanations"?
My idea was to e.g. use the segmentation masks (assuming we have one…
-
Hi, I was wondering if there is any way to use classes like the LimeImageExplainer or something similar in InterpretML?
-
I am trying to use `LayerDeepLift` on multiple layers of a VGG16 model from `torchvision.models`. It works for all layers except `MaxPooling2D` layers.
The following (layer `23` is a `MaxPool2d` la…
-
It seems to me the input in this example for visualize_image_attr_multiple is not the same as the argument format for the function `captum.attr.visualization.visualize_image_attr_multiple(attr, origin…
-
When displaying the attribution, you normalise and [scale](https://github.com/pytorch/captum/blob/28b89258163827b1bce7fb3543aae7d74b0fb4a0/captum/attr/_utils/visualization.py#L43) the values.
Howe…
-
Got a `ValueError: not enough values to unpack (expected 2, got 0)` error on `test, _ = test.split(split_ratio = 0.04)`.
Package version:
captum==0.2.0
spacy==2.3.0
torchtext==0.6.0
And:
en-cor…
-
Hi, first of all thanks for the amazing work you have done with Captum!
I have a question on the Deconvolution implementation. Most likely I'm missing an important implementation detail or using th…
-
## ❓ Questions and Help
Although I changed the width, the size of the plot is very small (width > 20 will not increase the plot anymore).
Also, as this function return `None`, how can I save the…
-
I am trying to use [HuggingFace's XLNet](https://huggingface.co/transformers/model_doc/xlnet.html) ```word_embedding```layer in LayerIntegratedGradients to calculate attributions. The method I use is…
kh8fb updated
4 years ago
-
## 🚀 Feature
According to the feedback we had from various sources:
> For me it took time to really understand the advantages of using the package. The docs i read were explanatory and didn't ta…