ReCoN-Ipsundrum: An Inspectable Recurrent Persistence Loop Agent with Affect-Coupled Control and Mechanism-Linked Consciousness Indicator Assays

Researchers developed ReCoN-Ipsundrum, an AI agent architecture designed to exhibit consciousness-like behaviors through recurrent persistence loops and affect-coupled control mechanisms. The study demonstrates how engineered systems can display preference stability, exploratory scanning, and sustained caution behaviors that mimic aspects of conscious experience.

• →ReCoN-Ipsundrum implements consciousness indicators through recurrent persistence loops and affect-coupled control systems.
• →Affect-coupled variants showed preference stability independent of novelty, unlike non-affect variants that remained novelty-sensitive.
• →The system demonstrated structured exploratory behavior with significantly higher scan events (31.4 vs 0.9) in reward-free environments.
• →Lesioning experiments confirmed causal links between specific mechanisms and consciousness-like behaviors.
• →The research provides a framework for engineering and validating consciousness indicators in AI systems.