method
active
method:edge-attribution-patchingEdge attribution patching
Used by prior work to identify consistency heads.
Neighborhood — ranked by edge-count
Papers (1)
paper
Methods (1)
method
- Attribution patchingrelated_toGradient-based method to estimate the effect of zeroing a feature on a specific logit difference.
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.
- Gradient-based approximation to activation patching.
- Gradient-based technique using SAE features to estimate causal effects on completions; used to corroborate NLA findings.
- Method by Goldowsky-Dill et al. 2023 for localizing model behavior via targeted activation interventions
- Measure causal importance of upstream variables.
- Baseline method against which probe-based ranking is compared; more computationally expensive.
- Intervention technique clamping activations prior to the intervention layer.
- Used by APD; poor approximations of causal importance.