Vanilla interchange intervention

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Interchange Interventionmethod0.843
  Fundamental operation for causal abstraction analysis; forces neurons to take values from source inputs to create counterfactuals.
• Distributed Interchange Interventionmethod0.779
  Extends interchange interventions to non-standard bases by rotating representations, intervening in rotated subspaces, then rotating back.
• Interchange Intervention Accuracymethod0.772
  Proportion of aligned interchange interventions with equivalent high-level and low-level effects; graded measure of causal abstraction.
• Interchange Intervention Training Objectivemethod0.760
  Differentiable training objective minimized when a high-level model is an abstraction of a neural network under a given alignment.
• Serial Interventionconcept0.735
  Intervention mode where interventions are applied sequentially, each building on the previous one
• Parallel Interventionconcept0.726
  Intervention mode where multiple interventions are applied simultaneously to the same base computation graph
• Intervention Propagationconcept0.725
  Property that additive modifications to activations affect all downstream computations, enabling tractable behavioral control
• Subspace Interventionconcept0.724
  Intervention targeting specific dimensional subsets of activation vectors rather than full representations