method
active
method:forking-paths-analysis-fpaForking Paths Analysis (FPA)
Resampling method for measuring uncertainty dynamics along a base path.
Neighborhood — ranked by edge-count
Papers (1)
paper
Frameworks (1)
framework
- Change point segmentation, kernel Dirichlet pooling, CV tuning.
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.
- The path in activation space derived by optimizing steering interventions to produce outputs along the behavior manifold, independent of representation geometry.
- Framework for comparing trade-offs between trait expression and coherency across steering locations
- 1000-iteration bootstrap procedure sampling 50% of conTest to compute 95% confidence intervals
- Strategy forcing edges to be mediated only by head paths of length 1 to avoid exponential path growth.
- Explains advantage over importance sampling.
- Method by Goldowsky-Dill et al. 2023 for localizing model behavior via targeted activation interventions
- The core analytical technique of expanding transformer computations from layer-by-layer products into sums of end-to-end path terms for independent analysis