Comparative Database of Biologically Inspired Cognitive Architectures for AI
The disconnect between scientific schools of thought across the world is most notable in cognitive, neural, and computer sciences. The intersection of these fields is exactly where a powerful new approach has emerged recently, known as Biologically Inspired Cognitive Architectures (BICA). Several years ago, the BICA Society started a new initiative: to build an international community of researchers unified by the BICA Challenge – to create a computational replica of the human mind using solutions inspired by the brain. One of the key elements of the roadmap is a public online repository developed by collective efforts. We call it the BICA Repository, named by the following six components: Models, Architectures, People, Paradigms, Evaluations, and Dialogues, defined as follows.
- Models are task-specific and are represented by research papers, code, archived presentations, reports, manuals, and other documentation.
- Architectures are task-independent and are formulated computationally, at any level: from algorithms and pseudocode to software packages, APIs, and online tools.
- People include researchers and developers who are related or potentially related to particular BICA modeling approaches.
- Paradigms include methodology, tests, challenges, decathlons, testbeds, simulation environments, settings and procedures, research questions, hypotheses and predictions, measures, scales, and metrics, measuring tools, evaluation protocols, statistical analysis methods, and visualization tools.
- Evaluations include empirical and analytic data resulting from tests, analyses, and comparisons of models to each other.
- Dialogues among research groups and schools of thought have been initiated and facilitated by the BICA Society and are archived in the repository in various forms, documenting peer-to-peer interactions among developer groups who disagree on terminology, interpretation of results, values and priorities for research, or are missing facts and knowledge that they want to learn from each other. Forms of dialogues include teleconferences, Wiki pages, webinars and Videopanels, discussion boards, and workshops.
Each BICA is mapped to these six components, and vice versa; therefore, the repository represents cognitive architectures and models in a universal format, facilitating comparison and learning of models. We encourage all BICA researchers to support us and contribute their materials and efforts to this initiative.
— Version 2: David J Kelley
First Step: Comparative Table of Implemented Cognitive Architectures
- Current comparative table: XLS | PDF
- (currently, the table includes input from all contributors as is)
- Download template to contribute your architecture via email to admin@ArtificialGeneralIntelligenceInc.com
- Old comparative table – from Pew & Mavor, 1998
Complementary Frameworks for Comparison
- Hierarchy of Levels of Cognition
- CogAff (by Aaron Sloman)
- A Framework for comparing agent architectures, by Aaron Sloman and Matthias Scheutz
- ConsScale: a biologically-inspired scale for measuring cognitive development in natural and artificial creatures (by Raul Arrabales, also includes rating method and examples of cognitive architectures)
- Genealogical tree of cognitive architectures (to be added)