concept
active
concept:orthogonal-decomposition-of-representation-space

Orthogonal Decomposition of Representation Space

Mathematical structure central to distributed interchange interventions; representation space decomposed into orthogonal subspaces each aligned with a high-level variable.

Neighborhood — ranked by edge-count

Methods (1)

method
  • Extends interchange interventions to non-standard bases by rotating representations, intervening in rotated subspaces, then rotating back.

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

  • Investigation of whether a distributed representation can be further decomposed into sub-representations encoding component identities.
  • The idea that features are encoded as directions in activation space.
  • Internal structure of AI systems that CIMC proposes to analyze interpretively to evaluate consciousness hypotheses
  • Key mathematical object in DAS; transforms neural representations to alternative bases to reveal distributed structure.
  • Idea that information is spread across many neurons; superposition is a subtype.
  • CIMC's characterization of part of the solution to the Hard Problem: insight into the structural necessities of phenomenal representation
  • Substrate on which causal emergence was computed across agent lifetimes; aligned with learning success.
  • The central question of whether representational geometry implies corresponding computational structure