Stability, optimal-control reductions, and objective-function distinctions in MEM

Establish the stability properties of the Motivated Emotional Mind (MEM) mathematical model, identify the conditions under which it reduces to optimal-control or motivated-learning schemes, and clarify how its regulatory objective function differs from a conventional reward function.

Background

The paper presents MEM as a formal framework integrating embodied perception, associative representations called semblions, allostatic regulation, affective modulation, curiosity, and program-based action selection. Its policy minimizes discounted regulatory costs arising from deviations of allostatic variables from tolerated ranges rather than maximizing an externally specified reward.

The authors explicitly identify three unresolved theoretical issues: determining whether the coupled representational, affective, and regulatory dynamics are stable; characterizing when the formulation is mathematically equivalent or reducible to optimal-control or motivated-learning approaches; and distinguishing the regulatory-cost objective from the reward objectives used in conventional reinforcement-learning formulations.