It’s Doctor Hargrave now
Mason defended today his thesis entitled Benchmarking Reinforcement Learning for Optimizing Longitudinal Healthcare. Please join us for the traditional celebration at the Faculty Club today at 6pm.
Mason defended today his thesis entitled Benchmarking Reinforcement Learning for Optimizing Longitudinal Healthcare. Please join us for the traditional celebration at the Faculty Club today at 6pm.
Brigid successfully defended her Ph. D. thesis today! Congratulations Dr. Maloney!
Abstract: Changing a circuit dynamically, without actually changing the hardware itself, is called reconfiguration, and is of great importance due to its manifold technological applications. Circuit reconfiguration appears to be a feature of the cerebral cortex, and hence understanding the neuroarchitectural and dynamical features underlying self-reconfiguration may prove key to elucidate brain function. We present a …
Read more “Input-driven circuit reconfiguration in critical recurrent neural networks”
We just uploaded to arXiv: arXiv:2405.13854 Authors:Aditi Chandra, Marcelo O. Magnasco Abstract: We examine the dynamical properties of a single-layer convolutional recurrent network with a smooth sigmoidal activation function, for small values of the inputs and when the convolution kernel is unitary, so all eigenvalues lie exactly at the unit circle. Such networks have a variety of …
Read more “On the dynamics of convolutional recurrent neural networks near their critical point”
Posted to biorxiv on May 12, 2023 Eric A. Ramos, Mariam Steinblatt, Rachel Demsey, Diana Reiss, View ORCID Profile Marcelo O. Magnasco doi: https://doi.org/10.1101/2023.05.11.540348 ABSTRACT This paper presents some unusual behaviors observed in one single specimen of O. insularis. While nothing can be concluded rigorously from such data, we share the data and our analysis with the community, in the hope that others will be on …
Posted to arXiv:2302.07396 on Feb 2023. Authors: Marcelo O. Magnasco Abstract: Recurrent neural networks are extremely powerful yet hard to train. One of their issues is the vanishing gradient problem, whereby propagation of training signals may be exponentially attenuated, freezing training. Use of orthogonal or unitary matrices, whose powers neither explode nor decay, has been proposed …
Read more “Convolutional unitary or orthogonal recurrent neural networks”