This preprint presents Entropy-Regulated Artificial Cognition (ERAC), a cooperative dual-regulation framework that extends optimization-centered artificial intelligence by introducing continuous internal organizational regulation. In the proposed architecture, external optimization remains responsible for capability improvement, while internal regulation maintains organizational coherence and cognitive stability during long-horizon recursive reasoning. The two processes operate cooperatively rather than competitively. This version is a condensed manuscript prepared for journal submission. Detailed experimental protocols, statistical analyses, and complete results are available in a companion Zenodo preprint (DOI: 10.5281/zenodo.21697048). This is the 8th preprint version of the ERAC framework.
YOUNG KYU LEE (Sun,) studied this question.