extrapolation
Latest
Perception during visual disruptions
Visual perception is perceived as continuous despite frequent disruptions in our visual environment. For example, internal events, such as saccadic eye-movements, and external events, such as object occlusion temporarily prevent visual information from reaching the brain. Combining evidence from these two models of visual disruption (occlusion and saccades), we will describe what information is maintained and how it is updated across the sensory interruption. Lina Teichmann will focus on dynamic occlusion and demonstrate how object motion is processed through perceptual gaps. Grace Edwards will then describe what pre-saccadic information is maintained across a saccade and how it interacts with post-saccadic processing in retinotopically relevant areas of the early visual cortex. Both occlusion and saccades provide a window into how the brain bridges perceptual disruptions. Our evidence thus far suggests a role for extrapolation, integration, and potentially suppression in both models. Combining evidence from these typically separate fields enables us to determine if there is a set of mechanisms which support visual processing during visual disruptions in general.
Parametric control of flexible timing through low-dimensional neural manifolds
Biological brains possess an exceptional ability to infer relevant behavioral responses to a wide range of stimuli from only a few examples. This capacity to generalize beyond the training set has been proven particularly challenging to realize in artificial systems. How neural processes enable this capacity to extrapolate to novel stimuli is a fundamental open question. A prominent but underexplored hypothesis suggests that generalization is facilitated by a low-dimensional organization of collective neural activity, yet evidence for the underlying neural mechanisms remains wanting. Combining network modeling, theory and neural data analysis, we tested this hypothesis in the framework of flexible timing tasks, which rely on the interplay between inputs and recurrent dynamics. We first trained recurrent neural networks on a set of timing tasks while minimizing the dimensionality of neural activity by imposing low-rank constraints on the connectivity, and compared the performance and generalization capabilities with networks trained without any constraint. We then examined the trained networks, characterized the dynamical mechanisms underlying the computations, and verified their predictions in neural recordings. Our key finding is that low-dimensional dynamics strongly increases the ability to extrapolate to inputs outside of the range used in training. Critically, this capacity to generalize relies on controlling the low-dimensional dynamics by a parametric contextual input. We found that this parametric control of extrapolation was based on a mechanism where tonic inputs modulate the dynamics along non-linear manifolds in activity space while preserving their geometry. Comparisons with neural recordings in the dorsomedial frontal cortex of macaque monkeys performing flexible timing tasks confirmed the geometric and dynamical signatures of this mechanism. Altogether, our results tie together a number of previous experimental findings and suggest that the low-dimensional organization of neural dynamics plays a central role in generalizable behaviors.
Modeling Decision-Making in Trajectory Extrapolation Tasks: Comparing Random Sampling Model and Multi-Layer Perceptron Approaches
Bernstein Conference 2024
Cortically motivated recurrence enables visual task extrapolation
COSYNE 2023
Mood as an Extrapolation Engine for Adaptive Learning \& Decision-Making
COSYNE 2025
extrapolation coverage
5 items
Share your knowledge
Know something about extrapolation? Help the community by contributing seminars, talks, or research.
Contribute content