Talk by Prof. Rubén Moreno Bote (CSN Virtual Seminar)

We hereby announce the next talk in the 'CSN Virtual Seminar' series

Seeking entropy: Complex behavior and neural dynamics from maximally occupying action-state path space

Speaker: Prof. Rubén Moreno Bote (Universitat Pompeu Fabra)

Abstract

Most theories of behavior posit that agents tend to maximize some form of reward or utility. However, animals very often move with curiosity and seem to be motivated in a reward-free manner. Here we abandon the idea of reward maximization, and propose that the sole goal of intelligent behavior is maximizing occupancy of future paths of actions and states, a principle that we call path occupancy maximization. According to this view, rewards are the means to occupy path space, not the goal per se; goal-directedness simply emerges as rational ways of searching for resources so that movement, understood amply, never ends. We find that action-state path entropy is the only measure consistent with additivity and other intuitive properties of expected future action-state path occupancy. We provide analytical expressions that relate the optimal policy and state-value function, and prove convergence of our value iteration algorithm. Using discrete and continuous state tasks, we show that complex behaviors such as `dancing', hide-and-seek and a basic form of altruistic behavior naturally result from the intrinsic motivation to occupy path space. We finally extend our theory to neural dynamics, and show that it is possible to control chaotic systems while generating large variability so that state-space is maximally occupied.

You can dial in using the following zoom room information:
https://zoom.us/j/99314955130?pwd=YXkvaW1wY2s2OHJsMFRWVDNtcERiUT09
Meeting ID: 993 1495 5130
Passcode: 4pM2Z8