🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

New version of AEStream is out! 🔥 Blazing fast, NVIDIA-optimized streaming of event-camera data (using iniVation and Prophesee 🎥) perfectly suitable for #neuromorphic computing and deep learning projects with #norseai Check it out, it's #FOSS 🥳 github.com/norse/aestream/

On my way to the second Norse (github.com/norse) users meeting at FZI in Karlsruhe. It feels really good, when other people start using what you co-created. I am looking forward to hear about all the use cases they have for spiking neural networks.

I've joined a mastodon instance: @[email protected]

I'm looking for a PostDoc / tenure-track position starting next year. I am broadly interested in both algorithmic and theoretical aspects of physical computing and in particular spiking neurons, neuromorphic and quantum computing. Any leads or suggestions are appreciated!

New preprint: Loss shaping enhances exact gradient learning with EventProp in Spiking Neural Networks arxiv.org/abs/2212.01232 with James Paul Turner and Jamie Knight. We tested the EventProp algorithm of Timo Wunderlich & Christian Pehle on the Spiking Heidelberg Digits 🧵 1/5

Analog neuromorphic hardware still needs lots of software to operate: for machine learning on BrainScaleS 2 we have designed an in-the-loop experiment flow that enables gradient-based optimization in PyTorch of the emulated spiking neural networks arxiv.org/abs/2212.12210

Spatial and temporal correlations in neural networks with structured connectivity, Yan-Liang Shi et al Yanliang Shi Roxana Zeraati Anna Levina Tatiana Engel #Neuroscience go.aps.org/3ipstkX More in our Collection on the Physics of Neuroscience: go.aps.org/3bCA0cm

Palacio de Cristal, Madrid.

Yann LeCun @seanescola Blake Richards @BOlveczky finally! Dmitri "Mitya" Anne Churchland ClopathLab James DiCarlo Surya Ganguli Konrad Kording ioio Timothy Lillicrap Adam Marblestone Alexandre Pouget Sara A Solla Terrence Sejnowski David Sussillo Andreas Tolias Lab @ Stanford University Doris Tsao rdcu.be/c8dvd

Think that neurons are feature detectors or efficient coders? I will present evidence that neurons are feedback controllers, stabilizing other neurons and external environment. Implications for Neuroscience and artificial networks of feedforward McCulloch-Pitts-Rosenblatt units.

Generative AI is still stuck in the world of bits. 👨‍💻🖥️ It's high time for us to build generative AI for the world of atoms. ⚛️🤖 On it. 🫡😉🔥

Excited to share that our work on Bayesian + game-theoretic modeling of real-world social behaviors was selected to be presented at #NeurIPS2023 . Looking forward to fruitful discussions in New Orleans!

Btw, it's accepted at NeurIPS 2023: Trial matching paper, how to fit a large spiking network to thousands of recorded neurons on a single trial basis. Beautiful code published and prepared by Christos Sourmpis : github.com/EPFL-LCN/pub-s…

Last chance to apply for a neuroAI summer internship at beautiful CSHL! Please retweet!

We need your help for TENSS to carry on and inspire the next generations of neuroscientists. Logistics costs have increased and funding sources are becoming scarcer. We are counting on support from the community. Please donate! gofund.me/a8e6ad34

I'm at NeurIPS for the week — come by our spotlight poster on the "Neural Tangent Ensemble" — we show how a single neural network can be interpreted as an ensemble and apply this perspective to the problem of continual learning — arxiv.org/abs/2408.17394.