Structured dynamics in the algorithmic agent
bioRxiv Preprint Server December 12, 2023 G. Ruffini, F. Castaldo, J. Vohryzek 3 citations preprint
The paper extends the Kolmogorov Theory of Consciousness by examining how agents that track natural data develop structural and dynamical properties mirroring the world's symmetries. Using Lie pseudogroups to formalize generative models and a neural network as a proxy for the agent, the authors show that data tracking forces the agent's parameters and dynamics to reflect the symmetry properties of the generative world model, leading to a hierarchical organization consistent with the manifold hypothesis. This bridges algorithmic information theory, symmetry, and dynamics, offering insights into neural correlates of agenthood and structured experience.