Tobias Pitters

@hardikyagnik, the structure for dataset entry for SODA would be ```python DatasetEntry( questions=[PREFIX + Q1, Q2, Q3, ...], answers=[A1, A2, A2, ...] ```

@sampatkalyan yes, that is correct, please verify that you match the types in the annotations

@sampatkalyan Please create the PR, I'll have a look.

If it's just a type annotation, put it in the same PR

Thanks for the report. We are currently using dmatrix properties for xgboost models, see e.g. [here](https://github.com/shap/shap/blob/87ee4efde9d159456f80f68d4bdafd16c22c2d1a/shap/explainers/_tree.py#L1720). Note that we are only passing specific dmatrix properties, namely: ```python properties_to_pass = ["missing",...

I cannot reproduce this. This works for me on the latest master.

@connortann Is this already solved by #3550?

No objections from my side. Let's go ahead.

> Done 🎉 Let me know if you have any proposed edits to the release notes, in particular to the description of the breaking change to the shape of Explanation...

Thanks for the report @yaba-hun. From the code you posted I cannot understand what is going wrong and in your example there are a few things missing for us to...