finding
active
finding:r-0-99-p-0-0001-correlation-between-features-per-dimension-and-adversarial-vulnerability-in-toy-models-with-elhage-style-attacksr≈0.99 (p<0.0001) correlation between features per dimension and adversarial vulnerability in toy models with Elhage-style attacks.
Appendix replication.
Source paper
extracted_from(2025) · Gorton, Liv · Lewis, Owen
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Central thesis.
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.
- Robustness controls superposition.
- Superposition controls robustness.
- Adversarial vulnerability increases >3x as superposition forms and tracks features per dimensionfinding0.812L2 attacks max norm 0.1 of average input norm.
- SAE features are not simply mirroring individual neurons.
- Summary finding of the full behavioral sweep
- Unsatisfying reduction.
- Qualitatively different defense profile compared to Llama-3.1-8B
- Architecture-specific difference in trait vector geometry