Synaptic consolidation across multiple timescales

Max-weight scheduling across multiple timescales

Validating Computational Cognitive Process Models across Multiple Timescales

Model comparison is vital to evaluating progress in the fields of artificial general intelligence (AGI) and cognitive architecture. As they mature, AGI and cognitive architectures will become increasingly capable of providing a single model that completes a multitude of tasks, some of which the model was not specifically engineered to perform. These models will be expected to operate for extend...

Synaptic Learning Rules with Consolidation

How do we remember the past ? By what means can we make sense of our environment and store its most relevant aspects ? Learning and memory is very important for the existence of complex behaviours in living animals since it is what enables the creation of an internal model of the world in order to take the best possible decisions. The theory of synaptic tagging and capture (STC) represents a po...

The multiple timescales of memory

A few years ago, Yoshiko Kojima, Yoshiki Iwamoto, and Kaoru Yoshida (2004) took their monkey to a darkened room and measured its eye movements (saccades) in response to display of visual targets. Usually, when a target is shown at 10 o with respect to current fixation, the animal makes a 10 o saccade to foveate the target. If we define saccadic gain as the ratio between the displacement of the ...

Unravelling cross-recurrence: coupling across timescales

We present an extension of cross-recurrence analysis for dual gaze analysis which is suited for complex situations where for instance the objects of interest are not all visible at all times or when stimulus exploration is not homogeneous. The typicical situation is a visual stimulus that is scrolled or that is explored sequentially. We use a recurrence simulation to illustrate how to measure t...