question
active
question:how-important-is-the-geometry-of-activation-space-for-explaining-neural-network-behaviorHow important is the geometry of activation space for explaining neural network behavior?
Open question A.1.1a.5.
Source paper
extracted_from(2025) · Lee Sharkey · Bilal I. Chughtai · Joshua Batson · Jack W. Lindsey +25
Neighborhood — ranked by edge-count
Papers (1)
paper
Claims (1)
claim
- Calls for conceptual clarity.
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.
- Does the geometric structure of activation space causally shape neural network behavior?question0.895Central research question driving the work.
- Linear representation hypothesis supported by empirical and theoretical arguments.
- Closing statement.
- The core testable hypothesis driving the experimental design
- Rich geometric structure carried by neural representations.
- The opening statement of the paper, framing concept geometry as the key to neural network control.
- Central empirical claim of the paper, demonstrated across tasks and modalities