Quantus

Quantus copied to clipboard

### Description - Here https://github.com/understandable-machine-intelligence-lab/Quantus/blob/main/quantus/helpers/model/tf_model.py#L272 for Keras model we shuffle each layer weights. - Here https://github.com/understandable-machine-intelligence-lab/Quantus/blob/main/quantus/helpers/model/pytorch_model.py#L213 for torch model we reset module parameters (to default ones.) ### Minimum acceptance criteria...

aaarrti

Missing module in 'quantus.helpers.utils'

2

A error just popped up which is required for torch models when I want to evaluate XAI methods: ---------------------------------------------------------------------------------------------------------- AttributeError Traceback (most recent call last) [](https://localhost:8080/#) in () 6 7...

PouyaKhn

Implement custom plot function

1

Hi! I want to plot the results of some Quantus metrics with different types of graphs for easy comparison when researching, how can I implement a custom plot function which...

binh-nguyen49-jh

### Description of the problem - Quantus provides examples of simple models to get started with, e.g. https://github.com/understandable-machine-intelligence-lab/Quantus/blob/main/quantus/helpers/model/models.py#L128 - It is however up to user to download the weights from...

aaarrti

### Description The .whl file uploaded to GitHub release, and the one on PyPI have different sha256 sums. Probably because, they were built on different hosts (mb event with different...

aaarrti

### Description of the problem - Write a short guide on how to use Quantus for different datasets (readthedocs.org), under `Getting Started` - https://quantus.readthedocs.io/en/latest/ ### Minimum acceptance criteria - Specify...

annahedstroem

### Description - Huggingface package ### Steps to reproduce the behavior - If installing quantus with conda Python 3.8, this package issue appear:  ### Minimum acceptance criteria - Specify...

annahedstroem

### Description of the problem & solution Currently, the docstring of each metric mostly explains what the metric computes, as well as the corresponding reference. It may be helpful in...

leanderweber

attribution localisation metric should consider positive attributions only

1

### Description According to the paper (https://arxiv.org/pdf/1910.09840.pdf) attribution localisation should be computed only from the positive values in the attribution map. The implementation in quantus considers negative values also in...

leanderweber

### Description of the problem - attribution preprocessing before metrics should be available as separate modules which could be pipelined, like torchvision.transform or sklearn Pipeline. - At the moment, attribution...

aaarrti