Low-Dimensional Attractor Dynamics in RNNs

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

• Random dynamical attractorconcept0.740
  The invariant set of states to which a random dynamical system converges.
• Attractor networks in morphogenesisconcept0.729
  The idea that bioelectric circuits form dynamical attractors corresponding to stable anatomical states, analogous to Hopfield networks.
• Random Global Attractorconcept0.725
  Invariant set of states toward which ergodic random dynamical systems converge over time.
• Emergence of grid-like representations by training recurrent neural networks to perform spatial localization (Cueva & Wei, 2018)concept0.712
  RNN model recapitulating grid cells; related work category 4.
• Attractorsconcept0.708
  Dynamical concept: stable regions in a system's state space; proximate intention becomes attractor that entrains motor dynamics.
• Attractor in State-Spaceconcept0.707
  Small regions in dynamical system state-space to which complex systems converge; cited from complexity theory as partial but insufficient explanation for living structure
• We hypothesize that a Representation Network (RN) emerges from LLM representations, where each dimension is a node and latent connections exist between nodes or clusters of nodes.hypothesis0.701
  Core methodological hypothesis enabling the application of IIT to LLM representation sequences.
• Attractor Stateconcept0.691
  Low-energy configuration toward which systems are drawn; low-stress states serve as attractors in morphogenesis.