Neuro-Symbolic AI
EDUCATION
What is
Neuro-Symbolic AI?
Neuro-symbolic AI refers to a loose family of approaches which each somehow incorporate “symbolic” ideas or inspirations, such as the KRR (knowledge representation and reasoning) techniques of logic-based “classical AI” and those of probabilistic reasoning, with the complementary ideas of neural networks. Within the umbrella are a number of distinct approaches which differ in some key aspects from each other, but in general all of the approaches are aimed at treating one or more of the weaknesses of neural networks while somehow maintaining their strengths to some degree.
Why can neuro-symbolic AI in principle all of the main safety and reliability issues of AI? see AG video.
The Challenge of
Learning
Neuro-Symbolic AI
Though many have long felt that many of the answers the field of AI is seeking has a good chance of lying in the general vicinity of what has now been called neuro-symbolic AI, it has significant barriers to entry.
One of the main obstacles to the development of neuro-symbolic ideas is the inherent difficulty in spanning two very different intellectual traditions, each of which is itself vast and deep - that of statistical AI or machine learning (ML), and that of symbolic AI or knowledge representation and reasoning (KRR) – most researchers are trained in one side but not the other. We hope to help onboard new researchers into this pursuit by providing educational entry points for researchers from both sides.
-
Neuro-symbolic AI systems span a spectrum ranging from those that simply combine separate "neuro" and "symbolic" systems while leaving each intact, which are the easiest to understand (and there are advantages to this kind of combination) to those that more fuse concepts from both sides in deeper ways, which can open up powerful new opportunities, but also many questions: What is the essence of what it means to be a 'neural' approach? What is the essnce of what it means to be a 'symbolic' approach? Can one formalism actually be both at the same time? What tend to be the strengths and limitations of each approach, and why? What are the systematic theoretical and empirical ways to characterize these strengths and limitatons? Which properties tend to trade off with each other, either formally or empirically? What are the deeper principles underlying each of these that can idneitfy and utilize to create entirely new kinds of models that achieve more of our desiderata, or at least better tradeoff points?
Another barrier to entry is the accelerating pace of new approaches, voices, and themes within neuro-symbolic AI. We approach these two barriers by maintaining a large network of experts at the top of their fields throughout all of the sub-disciplines involved so that we can present a curated selection that is refreshed each year, balancing ideological and technical coverage with the time cost of learning entirely new areas for practicing researchers.
Neuro-Symbolic AI
READING GROUP
We will be kicking off our Winter 2025 Neuro-Symbolic AI Cycle Reading Group, starting in February. The format is that each week a paper will be presented in a one-hour talk, either by the authors or a volunteer, followed by extended discussion. The paper will be announced at least a week in advance so that everyone has a chance to read the paper in advance. Sign up for the mailing list to receive the announcements.
Neuro-Symbolic AI
SUMMER
SCHOOLS
Want to quickly pick up some topics you don't know well, so you can get going with research? Our summer schools are curated crash courses designed to allow students and researchers to quickly boostrap their knowledge to be able to understand and perform neuro-symbolic research, tools, and perspectives. We hold a Summer School every year (since 2022) to help the community to keep abreast of the latest ideas in the field, and provide a professional certification of attendance upon request.
NSSS 4 (4th Neuro-Symbolic AI Summer School 2025)
Coming in August 2025. Sign up to receive the announcements.
NSSS 3 (3rd Neuro-Symbolic AI Summer School 2024)
Register for free to again access to the recordings.
AAAI Workshop (NucLeaR 2024)
Past Events
-
We have started developing a comprehensive course — join our mailing list (below) to stay informed.
-
Those who hang out with us for a while may start to learn our culture of learning, in which learning new areas is not just a necessary one-time struggle, but instead some we celebrate as a continuous way of being. We make a habit of mutual learning and teaching with a culture of "no dumb questions" and generosity in sharing one's knowledge with others - embracing and even seeking out the struggle, and helping others through it.
-
Go here.
-
Go here.