framework
active
framework:relu-output-toy-modelReLU output toy model
x'=ReLU(W^TWx+b), main model demonstrating superposition.
Neighborhood — ranked by edge-count
Papers (1)
paper
- Toy Models of Superpositionintroducesmentions
Methods (1)
method
- Visualizes feature interference matrix.
Concepts (1)
concept
- Thomson problemanalogous_toPoint-packing problem analogous to the toy model loss.
Hypotheses (1)
hypothesis
- Central hypothesis.
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.
- Toy model with privileged basis.
- Core demonstration.
- Without superposition.
- A model mapping outputs to expected inputs, used in motor control and perception for embodiment.
- Baseline SAE architecture.
- Open question.
- SAE variant trained on ResNet18.