Partial Information Decomposition (PID)

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

• ΦID (Integrated Information Decomposition)concept0.754
  Recently proposed method for estimating causal emergence; applied to RL agent latent-space representations.
• VPD (adVersarial Parameter Decomposition)concept0.753
  Core methodological framework introduced in this paper; decomposes weight matrices into rank-one interpretable subcomponents using adversarial ablations.
• partial introspectionconcept0.742
  The authors' characterization of genuine but limited introspective capability found only in early-layer injection regimes
• partial involutionsconcept0.732
  Fixed-point free partial injective functions used as simple reversible dynamical processes in Geometry of Interaction.
• Adversarial Parameter Decomposition (VPD)method0.732
  Core technique introduced in this paper for decomposing neural network weight matrices into mechanistically simple, interpretable rank-one subcomponents.
• VPD subcomponents are sparse, interpretable, and avoid feature splitting.claim0.730
  Assertion about the qualitative advantages of VPD's rank-one decomposition.
• Partial orders naturally represent partial information states in computation.hypothesis0.725
  Foundational hypothesis of Domain Theory: partial order structure (D, ⊑) captures information ordering without quantification.
• Elements of domains form partially ordered information states where d ⊑ e means 'e conveys at least as much information as d'.finding0.720
  Core intuition of Domain Theory: qualitative ordering of information states provides foundation for modeling computation without quantification.