skorch models raising The SHAP explanations do not sum up to the model's output
oegedijk opened this issue · comments
Oege Dijk commented
Seems related to this bug: shap/shap#3363
Can be avoided with passing shap_kwargs=dict(check_additivity=False)
to the explainer, but then you might get inaccurate shap values.
Added check_additivity=False param to the skorch tests for now.