Artificial Intelligence Is Unlocking the Secrets of Black Holes. A neural network trained with simulations of supermassive black holes has found that the one at the center of the Milky Way, Sagittarius A*, likely rotates at maximum speed.
https://zurl.co/azgAS
#ai #blackhole #neuralnet
#Neuralnet
🤯OMG, someone actually thinks converting a neural net to C is the key to unlocking the secrets of the universe—who knew swapping activation functions for logic gates would be the 🚀 to a 1,744× speedup? Clearly, this coder spent 25 minutes bragging about reinventing the wheel—next up, they'll try to prove water is wet! 💧🔧
https://slightknack.dev/blog/difflogic/ #neuralnet #Cconversion #speedup #techinnovation #codinghumor #unlocksecrets #HackerNews #ngated
Dunno if you heard of #BeamEyeTracker before. It's positioning itself next to #Tobii but uses just a webcam and #Opentrack, so probably just #Neuralnet tracker on steroids - but I'd still appreciate if you could chime in for a #Linux build at https://steamcommunity.com/app/2375780/discussions/0/4757578278664103829/ too as it seems to be read by the dev|s.
#LinuxGaming #GamingOnLinux #HeadTracker #HeadTracking #TrackIR
"One common criticism of using deep generative models in an artistic and creative context, is that they can only reproduce samples that fit the distribution of samples in the training set. However, by introducing deterministic controlled filters into the computation graph during inference, these models can be used to produce a large array of novel results."
https://www.mdpi.com/1099-4300/24/1/28
#networkbending #neuralnet
What are your beliefs on transhumanism? #transhumanism #neuralnet #elonmusk
-𝟘: The Power of Negative Zero: Datatype Customization for Quantized Large Language Models
(... they get an amazing boost in performance from this!)
🧠 Selecting a model for semantic search at Dropbox scale
https://dropbox.tech/machine-learning/selecting-model-semantic-search-dropbox-ai
Ich finde solche Animationen jedes Mal sowas von futuristisch und high-tech 😳 Dass das mit unserer aktuellen Technologie möglich sein soll, ist immer noch ein wenig unglaublich 😲
https://huggingface.co/spaces/merle/PROTEIN_GENERATOR
Veröffentlichung dazu: https://www.nature.com/articles/s41587-024-02395-w
#ai #artificalintelligence actually #neuralnet #machinelearning
It has been a while that I tried #StarCitizen. With the new #Neuralnet Tracker plugin (AI haha) for #OpenTrack we get head tracking without annoying IR LEDs or reflecting stripes just by reading the webcam video feed. This is apparently fast enough to try #headTracking without a dedicated #headTracker nowadays. And all that on a #Linux PC. Took some fiddling but the concept still works. What a time to be alive.
Demo: https://makertube.net/w/groS1wpAhP8XYE75vJwX32
HowTo: https://simpit.dev/systems/opentrack/
https://beko.famkos.net/2024/09/02/15629325/
#gaming #headtracker #Headtracking #linux #Neuralnet #opentrack #StarCitizen
A New Type of Neural Network Is More Interpretable
https://zurl.co/oegu
#ai #neuralnet
Updated https://simpit.dev/ - home of the Space Pew Pew #SimPit inspired by a VF-1 Valkyrie cockpit (#Macross)
That's a glorified #DIY joystick controller with an LCD ('MFD') and plenty of RGB for my #Linux PC
Newest additions are my videos for #X4Foundations, #EliteDangerous and #FlightOfNova (all also on https://makertube.net/a/bekopharm) and instructions how to compile #OpenTrack with the #neuralnet tracker using #ONNX runtime.
Best viewed WITH an ad-blocker and patience (slow).
OpenAI backed 1X EVE humanoid robots demonstrate new neural network
https://zurl.co/jeuH
#ai #robotics #neuralnet
@scottmiller42 @emovulcan @acute_distress @thomasfuchs
Hey Scott, I have not decided what to think about LLMs and copyrights.
I saw the word Data analyst in your Bio, so I suppose you have some knowledge in #DataScience and #DataEngineering.
I am myself trained in #MachineLearning and a bit in #neuralnetworks but I am far far far far far away from being any close to an expert. I can use them, efficiently sometimes. Sometimes not: there are problems I am still bad at solving with them.
I am still learning the Transformers stuffs and other types of layers used in Deep Learning. That's huge stuff. For me at least.
Could you clear that doubt I have when I read your messages:
when you talk about "MEMORIZING" data
It conflicts with my poor understanding of #machinelearning #neuralnet #transformer
At my level of understanding: memorizing == overfitting
that means that the model cannot extrapolate or even interpolate between values of the training set.
This is something that I have categorized as a real flaw. If any of my models overfits, I consider it broken. I throw it away, it will not perform good on any unexpected input.
Could you explain me how "memorizing" would differ in your vision?
Don't hesitate to simplify things, I am not an expert yet at #transformer and #LLM
In advance thanks. I would not like to be fooled by my apriori.
[AI Coffee Break with Letitia] explains the #transformer architecture behind #LargeLanguageModels
https://youtu.be/ec9IQMiJBhs
#LLM #NeuralNet #ArtificialIntelligence
See and hear how an AI learns the shape of sinus curves: https://youtu.be/0T46pTfCSsE #ai #ki #neuralnet #sinus
AI ‘breakthrough’: #neuralnet has human-like ability to generalize language.
A #neuralnetwork based #artificialintelligence outperforms #ChatGPT at quickly folding new words into its #lexicon, a key aspect of human intelligence.
https://www.nature.com/articles/d41586-023-03272-3
#AI #generativeAI #machine_learning #machineLearning #SystematicGeneralization #SystematicGeneralisation
Where might I find the most accurate breakdown of all the sources of text used to train GPT-4?
That time I tried to evolve a set of neural networks to control Chicago Traffic and the damned thing cheated.
Client Info
Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst