peal.visualization

Rendering helpers the pipeline uses, plus the DiDAE table-1 script.

image_grid and model_comparison are imported by the pipeline itself (CFKD’s collages and the LRP comparison), visualize_counterfactual_gradients by the counterfactual explainer, and create_didae_table1 reads the result yamls a DiDAE run leaves under $PEAL_RUNS into the paper’s table 1.

Submodules

create_didae_table1	Build the DiDAE paper's Table 1 (LaTeX) from TensorBoard logs of CFKD runs.
image_grid	PIL/NumPy helpers that assemble PEAL's annotated image grids.
model_comparison	Side-by-side comparison grids of several models on the same samples.
visualize_counterfactual_gradients	Debug grids for gradient-based counterfactual search.