#Artifical_Intelligence प्रणाली को और सक्षम बना सकता है नया कृत्रिम न्यूरॉन #innovation #technology #hardware #Cyber #RAM Dr Jitendra Singh DSTIndia Ashutosh Sharma Principal Scientific Adviser, Govt. of India IIT Delhi V. Ramgopal Rao, Ph.D. Ministry of Education Technology Development Board Vigyan Prasar Dr Umashankar Mishra vigyanprasar.gov.in/isw/New-artifi…

Complements to IIT Delhi . We need to create commercial applications around this technology.

Thank you India Science Wire Vigyan Prasar for covering our r&d from #iitd supported by DSTIndia and SERB #makeinindia vigyanprasar.gov.in/isw/New-artifi…

Reading about the history, present, and future of #neuromorphic engineering is fascinating.

Dear neurons, Why aren't chemical synapses, electrical synapses, diffusible gasses, and ephaptic signalling enough for you? Now we have to think about physical force, too? Signed, Neuroscientists nature.com/articles/d4158…

The 70th anniversary of the Hodgkin-Huxley model is being celebrated at the Journal of Physiology, where it was published. On this occasion, there's a call for computational neuroscience manuscripts in a special issue on the subject. Cheers! physoc.onlinelibrary.wiley.com/hub/journal/14… #berglabcph

What questions do you want answered in today's debate? Reply / share this post. You can still sign-up, but spaces are limited: ow.ly/Mi4N50I48hX Yann LeCun Kwabena Boahen Giacomo Indiveri | @[email protected] AI at Meta NYU Center for Data Science NVIDIA AI Stanford Engineering Stanford Press Johns Hopkins Engineering Johns Hopkins University

There is a beauty to the biological substrate

"No pressure, no diamonds" is used often to justify massive workloads during learning. It should be "no rigor, no diamonds". Pressure is unhealthy. Rigor is not. #sciencezen

"Low Power Neuromorphic EMG Gesture Classification", Sai Sukruth Bezugam, Ahmed Shaban, Manan Suri arxiv.org/abs/2206.02061

Low Power Neuromorphic EMG Gesture Classification deepai.org/publication/lo… by Sai Sukruth Bezugam et al. including Ahmed Shaban #NeuralNetwork #Neurons

The next big breakthrough in AI will come from hardware, not software. Training giant models like PaLM already require 1000s of chips consuming several MW, and we will probably want to keep scaling these up several orders of magnitude. How can we do it? a 🧵

Low Power Neuromorphic EMG Gesture Classification buff.ly/3xByPBC #DL #AI #ML #DeepLearning #ArtificialIntelligence #MachineLearning #ComputerVision #AutonomousVehicles #NeuroMorphic #Robotics

As someone who works in both fields, the desire to compare supercomputers to brains always strikes me as weird. It is like comparing an office building to a redwood tree...

#FairUCNow

#IEDM coverage in Forbes (12/9): Reporting on emerging non-volatile memory technology papers from CEA-Leti, Imec & Samsung-- forbes.com/sites/tomcough…

Excited to be featured in the The New Indian Express for my work on #Neuromorphic #Computing, exploring more efficient and sustainable #AI systems. #hyderabad #iitd #ucsb #Sustainability #EnergyEfficiency #Research #Science" newindianexpress.com/cities/hyderab…

Annoyed to see papers taking model X (ResNet, Transformer) and replacing ReLU with a pseudo-Spiking activation (with a LIF for loop) 😩 Not good for computer science, never better than quantization Not good for neuroscience, time dimension makes no sense Do I miss something?

Like athletes adjusting speed mid-sprint, neurons adjust signaling for efficiency. Our review on Spike Frequency Adaptation Communications Engineering shows its potential to make Neuromorphic applications more energy-efficient, exploring challenges and opportunities. #AI #neuroscience

We made a blog breaking down concept of Spike Frequency Adaptation and other concepts discussed in the paper. Could be an interesting read. go.nature.com/3OIanaM