This project was developed at the IAS Research Centre for Life, Mind and Society at the University of the Basque Country, funded by a postdoc grant from this university.
One of the grand challenges for embodied cognitive science and neuroscience is to understand how behavior arises from the interaction of an organism’s nervous system, its body, and its environment. A limitation faced by embodied approaches is the lack of scalability of most of available analyses and modelling techniques. This limitation contrasts with recent successes in related areas using deep learning and predictive processing architectures, often using tools compatible with embodied approaches although with incompatible conceptual assumptions. The current project aims to explore how to circumvent some of these limitations by the combination of datadriven maximum entropy models –statistical mechanics models mapping the statistics of experimental data – with information theoretical tools exploiting the statistical descriptions of the former to provide better indicators of adaptive and embodied neural activity. The project combines theoretical analysis of how statistical models may exploit concepts from statistical thermodynamics and information theory in order to display adaptive behavior, with analysis of experimental models inferred from real world neural systems data to apply previous theoretical exploration
A description of the project can be found here.
Aguilera, M & Di Paolo, EA (2019). Integrated information in the thermodynamic limit. Neural Networks, Volume 114, pp 136-146. doi:10.1016/j.neunet.2019.03.001
Aguilera, M. (2019). Scaling Behaviour and Critical Phase Transitions in Integrated Information Theory. Entropy, 21(12), 1198. doi:10.3390/e21121198
Aguilera, M & Bedia, MG (2018). Adaptation to criticality through organizational invariance in embodied agents. Scientific Reports volume 8, Article number: 7723 (2018). doi:10.1038/s41598-018-25925-4
Aguilera, M & Bedia, MG (2018). Exploring Criticality as a Generic Adaptive Mechanism. Frontiers in Neurorobotics 12: 55. doi:10.3389/fnbot.2018.00055