The STAR Cognitive Cycle in Context: A Comparative Analysis of 44 Cognitive Architectures

Iuliia Kotseruba, John K. Tsotsos

Keywords: cognitive cycle, comparative analysis, Core Model, STAR architecture

TL;DR: We identify cognitive cycles, modules and connections of 43 architectures and aggregate them using Core Model framework. We then highlight understudied areas and discuss how we plan to address these gaps in our STAR architecture.

Abstract: This paper discusses the cognitive cycle of the Selective Tuning Attentive Reference (STAR) architecture and compares it to cognitive cycles of existing cognitive architectures at different levels of abstraction. First, we briefly discuss the purpose of the cognitive cycle and how it can be specified. Second, we propose a Core Model that includes modules and connections found in typical cognitive architectures. Third, we analyze descriptions of cognitive cycles of 43 other cognitive architectures and express them in the Core Model format to highlight existing gaps in research, such as perception, attention, learning from experience, and metacognition. A report on the current status of STAR is then provided. Lastly, we apply this framework to STAR cognitive cycle and discuss how it addresses some of the identified gaps.

Paper Track: Commentary

Submission Number: 2