For the complete documentation index, see llms.txt. This page is also available as Markdown.
Model Comparison
Causal models evolve. You might train a Digital Twin on January data, then retrain on February data. Or experiment with different configurations — omitting certain variables, adding domain knowledge, changing temporal settings.
Model Comparison lets you put two Digital Twin versions side by side and understand exactly what changed: which relationships appeared, disappeared, or reversed direction, and how the underlying probability distributions shifted.
This is essential for model governance. Before promoting a new model version, you need evidence of how it differs from the current one.
Accessing Model Comparison
Model Comparison is its own page, not a tab on the twin: the Compare page under the Digital Twins section of your workspace (at …/twins/compare). The twin and version selectors are in the upper right corner.
Selecting models
The comparison tool lets you compare any two Digital Twin versions — including versions from different twins.
Common scenarios:
Same twin, different versions — track how the model evolved over retraining
Same data, different configurations — compare the effect of constraints or excluded variables
Different time periods — see how causal relationships changed between datasets
Comparing twins
Once two models are selected, the comparison view shows both causal graphs side by side with structural differences highlighted.
Two versions compared in the Separate view. The Comparison Overview reports size, structural overlap, and evaluation for both models — here v1.1.0 carries 32 relationships to v1.0.0's 31, sharing 28 edges.
Switching to Diff merges both graphs into a single view, color-coding the edges unique to either model and any whose direction changed. The Comparison Overview panel summarizes the structural overlap, evaluation, equations, and parameters, with Details links into a full breakdown of each.
The merged diff view. The 28 shared edges render grey; the coloured edges are unique to one model or flipped direction between them — exactly the differences re-discovery introduced.
Interpreting differences
Structural changes
New edges — the model found a relationship not present before; may reflect new patterns in updated data
Removed edges — a previously discovered relationship is no longer supported; may indicate data drift
Direction changes — causality now flows the opposite way; investigate why
Parameter changes
Even when structure is identical, parameters can differ. The conditional influence between variables may have become stronger, weaker, or changed in functional form.
Questions to ask
Are structural changes expected given the data differences?
Do direction changes make domain sense?
Would these changes affect downstream decisions?
Creating a second twin to compare
To compare two configurations or time periods, create a second Digital Twin from the same (or a different) Data View. Both twins will then be available in the Compare page selectors.
