concept
active
concept:simplicity-of-transcoder-explanationsSimplicity (of transcoder explanations)
Second category of SAE/transcoder concerns, distinct from faithfulness.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Refactoring of concerns into mechanistic faithfulness and simplicity.
Concepts (1)
concept
- Mechanistic faithfulnessassociated_withIdea that decomposed components must causally reflect the model's computation; enforced by adversarial ablation in VPD.
Related by similarity (8)
cosine ≥ 0.65 · no typed edgeEntities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.
- Uses activations from one layer to predict the next; special case of crosscoders.
- Used to express MLP in monosemantic features, 4096 features with JumpReLU.
- Decomposition method for activations; VPD is compared against transcoders in sparsity-reconstruction tradeoff.
- Discussion.
- Mask virtual weights during training targeting interference.
- A transformation that removes unwanted centers and unnecessary complexity throughout the structure, leaving only essential centers.
- Toy computation y=abs(x) implemented by two ReLU neurons, used to study unfaithfulness.
- The chapter's foundational question.