Visualize: Figure 1 & 2 & 3 & 4.
Description: "Reasoning Chain"
"Our datasets contain annotated (valid and invalid) reasoning chains in support of an answer, allowing explanation classifier models to be trained and applied. We also find that using a variabilized version of the chains improves the models’ robustness."