Nio(@NioStrats) 's Twitter Profile Photo

2 missing piece QinLihong maybe mentioning if the other part of $NIO entire full stack architecture.

mentioned
1-onboard data processing
(AD chip/LiDAR/sensors/camera)compressed to 1 gateway

2-NTP3.0 telecommunications 5 domain controllers
3-NeuralNets model training

2 missing piece QinLihong maybe mentioning if the other part of $NIO #NIO entire full stack architecture.

mentioned
1-onboard data processing
(AD chip/LiDAR/sensors/camera)compressed to  1 gateway

2-NTP3.0 telecommunications 5 domain controllers
3-NeuralNets model training
account_circle
Jonathan Ochmann(@MON0KEE) 's Twitter Profile Photo

I won't port machine learning stuff I wrote for match.color.io to new engine in app.color.io. Match generates with (black box), Color.io will use to generate editable, reinforceable app state (not a black box).

account_circle
Sandeep Reddy(@docsunny100) 's Twitter Profile Photo


Beware the neural network bandwagon. Computing costs, harms, lack of explainability, and divergence from biology suggest we need a mix of techniques for AI, not just bigger networks. The brain works
differently!

#artificialintelligence #neuralnets #bubble #justsaying
Beware the neural network bandwagon. Computing costs, harms, lack of explainability, and divergence from biology suggest we need a mix of techniques for AI, not just bigger networks. The brain works
differently!
account_circle
Antonia Statt(@StattAntonia) 's Twitter Profile Photo

The newest paper of my amazing grad (Rusen Argun) and undergrad (Yu Fu) student is now on arXiv: arxiv.org/abs/2402.12199 We look at how we can bypass costly distance calculations in MD simulations by using NeuralNets.

The newest paper of my amazing grad (Rusen Argun) and undergrad (Yu Fu) student is now on arXiv: arxiv.org/abs/2402.12199 We look at how we can bypass costly distance calculations in MD simulations by using NeuralNets.
account_circle
Michael Pyrcz🌻(@GeostatsGuy) 's Twitter Profile Photo

To help my students learn , Honggeun Jo & I coded a very simple demo to classify by the length of randomly positioned lines.

Demonstrates scale and location invariance!

To dive deeper, we made functions to visualize the kernels and…

To help my students learn #deeplearning #convolutional #NeuralNets, @HonggeunJ & I coded a very simple demo #CNN to classify by the length of randomly positioned lines.

Demonstrates scale and location invariance!

To dive deeper, we made functions to visualize the kernels and…
account_circle
TechPodGabe(@TechPodGabe) 's Twitter Profile Photo

Can machine learning ever understand laws of the physical universe (statistics, thermodynamics, gravity) and use that knowledge to guide how it learns new things? I better articulate this question in this video: tiktok.com/t/ZT8fnDXtu/

account_circle
Lana Dominkovic(@lana_dominkovic) 's Twitter Profile Photo

Writing deep learning papers in latex and missing a way to represent neural net?

Check 💡💡 Latex code for drawing neural networks for reports and presentation??? ->>> PlotNeuralNet

github.com/HarisIqbal88/P…

Writing deep learning papers in latex and missing a way to represent neural net?

 Check 💡💡 Latex code for drawing neural networks for reports and presentation??? ->>> PlotNeuralNet 

github.com/HarisIqbal88/P…

#neuralnets #latex #Researchpaper #researchwriting #plotneuralnet
account_circle
Sevak Avakians(@SevakAvakians) 's Twitter Profile Photo

Still one of my favorite XKCD comics, perfectly describing methodology using . There is a better way. is an analytical AI/ML/R technology that conforms to and requirements.

bit.ly/3OEPbTB

Still one of my favorite XKCD comics, perfectly describing #machinelearning methodology using #neuralnets. There is a better way. #GAIuS is an analytical AI/ML/R technology that conforms to #responsbleAI and #ExCITE requirements.

bit.ly/3OEPbTB
account_circle