victordibia / neuralqa

NeuralQA: A Usable Library for Question Answering on Large Datasets with BERT
https://victordibia.github.io/neuralqa/
MIT License
231 stars 32 forks source link

Improve Explanations: Better explanation visualization, more Explanation methods #35

Open victordibia opened 4 years ago

victordibia commented 4 years ago

Improved Explanation Visualization

Currently, we use a simple color density approach to visualize importance. Early feedback suggests this is helpful for the user immediately see the most important words/tokens, but does not offer quantities or further interaction (e.g. top n). Can we make this better or provide better alternatives?

More Explanation Methods

Currently, explanations are based on vanilla gradients. We might want to explore:

Useful Resources