claim
active
claim:toy-model-predictions-are-consistent-with-empirical-observations-on-polysemanticity-in-real-networksToy model predictions are consistent with empirical observations on polysemanticity in real networks
Validation; geometry and dynamics less certain.
Neighborhood — ranked by edge-count
Papers (1)
paper
- Toy Models of Superpositionmentions
Findings (2)
finding
- Neuron-level phase change.
- Mirrors real networks.
Questions (1)
question
- Open question.
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.
- Unexpected result.
- Validates toy model.
- Figure 4.
- Prediction orthogonality thesis.
- Authors' hedged interpretation.
- Symmetry is not explained by the base toy model.
- Main contribution for interpretability.
- Probe predictions are systematically miscalibrated until isotonic regression corrects biasfinding0.757Calibration necessity.