Frontiers | Learned adaptive properties for mitigation of weight perturbations in embedded spiking networks Recent years have seen an increased importance of neural network inference in edge-based scenarios, which impose size and power constraints requiring novel c...

Full open access paper: www.frontiersin.org/journals/neu...

New paper is out! Analog neural networks have the potential for highly efficient computation at the edge, but show decreasing accuracy with sources of noise. We show that biologically motivated “context” enables recurrent neural networks to adaptively change their dynamics to mitigate these effects.

To apply, or for more information, please navigate to
www.sandia.gov/careers/
and search for one of the following three openings, according to your research interests and expertise:

- Neuromorphic Codesign: 696457
- Computer Science and AI Theory: 696476
- Computational Neuroscience: 696477

How will you impact the world around you? How will you impact the world around you? Start your career Join Our Talent Community Access Existing Account/Create New Account in PeopleSoft Why Sandia? Your work can contribute to the security, pea...

The team has a newly awarded 3-year project to investigate the interplay of biologically inspired and generative artificial intelligence algorithms with neuromorphic platforms. We welcome both recent graduates, and those who are nearing their thesis defense.

Sandia Neuroscience Sandia Neuroscience in the News! Sandia's Billion Neuron Neuromorphic System Read More Frances Chance's Dragonfly TED Talk Watch the video Brad Aimone presents for the NIH Wednesday Afternoon Lecture ...

The Department of Cognitive and Emerging Computing at Sandia National Laboratories (neuroscience.sandia.gov)
is excited to announce we are searching for multiple postdoctoral researchers, broadly in the area of neuromorphic engineering and computational neuroscience.

Chand lab is at SfN! (And I’m looking for postdocs on cells types and dynamics; reach out if you want to chat!)