vraven.darkside · function
analyse_activation_mediation
analyse_activation_mediation(model: 'torch.nn.Module', clean_inputs: 'Any', corrupted_inputs: 'Any', *, target: 'int | None', layer_names: 'Sequence[str]', include_reverse_noising: 'bool' = True, config: 'VRavenConfig | None' = None) -> 'AnalysisResult'Patch clean activations into a corrupted run and measure score recovery.
This is an intervention-specific mediation diagnostic. Recovery can contain
interaction effects and must not be interpreted as a pure natural indirect
effect without stronger causal assumptions.
Usage
from vraven.darkside import analyse_activation_mediation
result = analyse_activation_mediation(... )Source of truthThis page is generated from the checked-out public symbol, signature, and docstring. Pin the package version for reproducible work.