Grace Lindsay | Discrete Symbols vs Continuous Neurons | NUMEROUS NUMEROSITY 2021
Plenary session kindly contributed by Grace Lindsay in SEMF's 2021 Numerous Numerosity: https://semf.org.es/numerosity/
SESSION ABSTRACT
Artificial neural networks are becoming increasingly popular models of how the brain processes information. They've been shown capable of playing games at human level, predicting neural activity in response to real-world images, and capturing basic dynamics of decision making. In the majority of these networks, individual neurons can take any non-negative real value. Yet in the history of cognitive science, discrete and symbolic processing has been highlighted as a way of describing the mind. I will compare and contrast these views, explaining why continuous values are necessary for neural networks and how people are aiming to connect these two seemingly disparate approaches. Questions will include: What counts as a symbol? How can symbolic/discrete processing arise from continuous neurons? Will different or hybrid structures ultimately be needed to model the mind/brain? Do we need to reconcile these views or is the success of neural networks enough to support them as models on their own?
SESSION MATERIALS
· Letting Structure Emerge: Connectionist and Dynamical Systems Approaches to Cognition (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3056446)
· Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future (https://arxiv.org/abs/2001.07092)
· On the Binding Problem in Artificial Neural Networks (https://arxiv.org/abs/2012.05208)
· Symbolic Behaviour in Artificial Intelligence (https://arxiv.org/abs/2102.03406)
GRACE LINDSAY
Center for Theoretical Neuroscience (Columbia University): https://ctn.zuckermaninstitute.columbia.edu/
Personal website: https://gracewlindsay.com/
Twitter: https://twitter.com/neurograce?lang=es
LinkedIn: https://www.linkedin.com/in/grace-lindsay-40454126/?originalSubdomain=uk
SEMF NETWORKS
Website: https://semf.org.es
Twitter: https://twitter.com/semf_nexus
LinkedIn: https://www.linkedin.com/company/semf-nexus
Instagram: https://www.instagram.com/semf.nexus
Facebook: https://www.facebook.com/semf.nexus
Видео Grace Lindsay | Discrete Symbols vs Continuous Neurons | NUMEROUS NUMEROSITY 2021 канала SEMF
SESSION ABSTRACT
Artificial neural networks are becoming increasingly popular models of how the brain processes information. They've been shown capable of playing games at human level, predicting neural activity in response to real-world images, and capturing basic dynamics of decision making. In the majority of these networks, individual neurons can take any non-negative real value. Yet in the history of cognitive science, discrete and symbolic processing has been highlighted as a way of describing the mind. I will compare and contrast these views, explaining why continuous values are necessary for neural networks and how people are aiming to connect these two seemingly disparate approaches. Questions will include: What counts as a symbol? How can symbolic/discrete processing arise from continuous neurons? Will different or hybrid structures ultimately be needed to model the mind/brain? Do we need to reconcile these views or is the success of neural networks enough to support them as models on their own?
SESSION MATERIALS
· Letting Structure Emerge: Connectionist and Dynamical Systems Approaches to Cognition (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3056446)
· Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future (https://arxiv.org/abs/2001.07092)
· On the Binding Problem in Artificial Neural Networks (https://arxiv.org/abs/2012.05208)
· Symbolic Behaviour in Artificial Intelligence (https://arxiv.org/abs/2102.03406)
GRACE LINDSAY
Center for Theoretical Neuroscience (Columbia University): https://ctn.zuckermaninstitute.columbia.edu/
Personal website: https://gracewlindsay.com/
Twitter: https://twitter.com/neurograce?lang=es
LinkedIn: https://www.linkedin.com/in/grace-lindsay-40454126/?originalSubdomain=uk
SEMF NETWORKS
Website: https://semf.org.es
Twitter: https://twitter.com/semf_nexus
LinkedIn: https://www.linkedin.com/company/semf-nexus
Instagram: https://www.instagram.com/semf.nexus
Facebook: https://www.facebook.com/semf.nexus
Видео Grace Lindsay | Discrete Symbols vs Continuous Neurons | NUMEROUS NUMEROSITY 2021 канала SEMF
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Transdisciplinary Mathematics (5/5) | Carlos Zapata | Summer School 2022SEMFILOQUIO #2 | Filosofía | ¿Es la Filosofía para la Ciencia como la Ornitología para los Pájaros?SEMFILOQUIUM #1 | Art vs Science | Maximillian Schich, Carlos Zapata-CarrataláGrace Lindsay | Discrete Symbols vs Continuous Neurons | NUMEROUS NUMEROSITY 2021Brain Dynamics & Computation | Joana Soldado-Magraner | Summer School 2022SEMFILOQUIUM #3 | Foundations of Finance | Miroslav Kiš, Carlos Zapata-Carratala¿Qué Significa Calcular? | Carlos Zapata CarrataláMaterial Origin of Numbers | Karenleigh OvermannFunctional Analysis (6/8) | Óscar Roldán | Summer School 2022Is Our Number System Indeterminate? | Pablo Dopico | Numerosity Workshop 2021Simposio de Invierno 2020 | RetrospectivaInformation Theory of Individuality | David KrakauerDoes Science Need Infinities? | Menachem MagidorIn-Person Workshop | Into the Information Continuum🚀🥳Fun Raising🥳🚀 | Funding & School Q&A + SEMF SocialInterdisciplinary Summer School 2023 | WolframDeep Learning Crash-Course in 5 Minutes | Grace LindsayNeuromorphic Computing | Saray Soldado-Magraner | Summer School 2022Community Livestream | BioelectricityCommunity Livestream | The Limits of ScienceA Landscape of Cognitiveness | SEMF Summer School 2020