Redhotcyber · @redhotcyber
588 followers · 1792 posts · Server mastodon.bida.imLSTM e GRU: Miglioramenti delle Recurrent Neural Networks
Benvenuti al quarto articolo della nostra serie dedicata alle #Recurrent #Neural #Networks (#RNN).
#redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #cybersecurityawareness #cybersecuritytraining #cybersecuritynews #infosecurity
https://www.redhotcyber.com/post/lstm-e-gru-miglioramenti-delle-recurrent-neural-networks/
#recurrent #neural #networks #rnn #redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #CyberSecurityAwareness #cybersecuritytraining #CyberSecurityNews #infosecurity
Redhotcyber · @redhotcyber
558 followers · 1709 posts · Server mastodon.bida.im
Problemi delle Recurrent Neural Networks: Scomparsa del Gradiente
Benvenuti al terzo articolo della nostra serie dedicata alle reti neurali ricorrenti (#RNN). Dopo aver introdotto le #Recurrent #Neural #Networks e la loro #architettura, rispettivamente nel primo e nel secondo #articolo, in questo articolo affronteremo un problema comune che le affligge: la temibile scomparsa del #gradiente.
Condividi questo post se hai trovato la news interessante.
#redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #cybersecurityawareness #cybersecuritytraining #cybersecuritynews #infosecurity
https://www.redhotcyber.com/post/problemi-delle-recurrent-neural-networks-scomparsa-del-gradiente/
#rnn #recurrent #neural #networks #architettura #articolo #gradiente #redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #CyberSecurityAwareness #cybersecuritytraining #CyberSecurityNews #infosecurity
Redhotcyber · @redhotcyber
557 followers · 1695 posts · Server mastodon.bida.imIntroduzione alle Recurrent Neural Networks
Benvenuti a questa serie di articoli sulle #reti #neurali #ricorrenti (#RNN). Queste reti sono una parte cruciale del progresso nell’ambito dell’#intelligenza #artificiale e del #MachineLearning.
A cura di Simone Raponi del gruppo AI di RHC.
Condividi questo post se hai trovato la news interessante.
#redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #cybersecurityawareness #cybersecuritytraining #cybersecuritynews #infosecurity
https://www.redhotcyber.com/post/introduzione-alle-recurrent-neural-networks/
#reti #neurali #ricorrenti #rnn #intelligenza #artificiale #machinelearning #redhotcyber #online #it #web #ai #hacking #privacy #cybersecurity #cybercrime #intelligence #intelligenzaartificiale #informationsecurity #ethicalhacking #dataprotection #CyberSecurityAwareness #cybersecuritytraining #CyberSecurityNews #infosecurity
Frank der Entspannende 🥾👟🇺🇦 · @DerEntspannende
434 followers · 3426 posts · Server digitalcourage.social#ÖPNV in #Rheinhessen: Ab September werden keine Fahrten der Linie 660 angezeigt, weder bei #RNN noch bei #Bahn. Das Servicetelefon von RNN sagt mir großzügig, ich solle später wieder anrufen.
Es ist eh schon schwierig genug, wenn ich mit ÖPNV bzw. Fernverkehr eine doppelte oder dreimal so lange Reisezeit einplanen muss. Wenn ich dann auch noch keine Planungsgrundlage für Verbindungen habe, reise ich halt mit dem Auto, leider. Immerhin muss ich dann bei Morgenterminen keine zusätzliche Übernachtung einplanen.
Thomas Gottfried · @thomasgottfried
53 followers · 539 posts · Server legal.social
DurstewitzLab · @DurstewitzLab
178 followers · 6 posts · Server mathstodon.xyz
Our #ICML2023 paper introduces an adaptive gradient re-scaling trick to fully resolve the exploding grad. problem when training recurrent neural networks (#RNN) on chaotic #DynamicalSystems. By this we were able reconstruct DS underlying even challenging empirical data (like EEG) incl. their attractors:
https://arxiv.org/abs/2306.04406
While many algorithms manage good short-term predictions, what has been tricky so far is capturing the long-term dynamics. If you know of any other #ML/ #AI / #DataScience algos which can achieve this on real world data, please let us know!
Fantastic work by Florian Hess, Manuel Brenner, & Zahra Monfared!
Graphical abstract below ...
#datascience #ai #ml #dynamicalsystems #rnn #icml2023
New Submissions to TMLR · @tmlrsub
180 followers · 563 posts · Server sigmoid.social
GraphPNAS: Learning Probabilistic Graph Generators for Neural Architecture Search
Albert Cardona · @albertcardona
1755 followers · 2280 posts · Server mathstodon.xyz
"We propose a solution inspired by biological neuron dynamics that makes the communication between RNN units sparse and discrete." https://openreview.net/forum?id=lJdOlWg8td
Now you got my attention – thanks for sharing!
By the way the event-based approach to computational efficiency without compromising performance seems Delbruck-inspired. Amusingly, in the paper the DVS128 Gesture Dataset is used ... very fitting indeed.
ct · @ctaylor
4 followers · 155 posts · Server mastodon.content.townA very nice tutorial/introduction to RNNs
https://www.youtube.com/watch?v=ySEx_Bqxvvo
#rnn #vanishinggradients #compsci #statistics #appliedmath #computerscience #datascience
#rnn #vanishinggradients #compsci #statistics #appliedmath #computerscience #datascience
mikeiavelli · @mikeiavelli
26 followers · 38 posts · Server mathstodon.xyz
Recurrent Memory Transformer (RMT) retains information across up to 2 million tokens (!) ⭐, significantly exceeding the largest input size reported for transformer models (64K tokens) and GPT-4's 32K tokens.
Paper: https://arxiv.org/abs/2304.11062
Code: https://github.com/booydar/t5-experiments/tree/scaling-report
#MachineLearning #ML
#NeuralNetworks #NN
#DeepLearning
#LLM #Transformers #RNN
#rnn #transformers #llm #deeplearning #nn #neuralnetworks #ml #machinelearning
smashmaster0045 · @smashmaster
6 followers · 137 posts · Server mastodon.gamedev.place
RT @BlinkDL_AI
Raven v8🐦14B to the moon🚀based on #RWKV (100% #RNN language model) 14B/7B/3B/1B Download: https://huggingface.co/BlinkDL/rwkv-4-raven And v9 soon (ctxlen 8192, 3x SFT data)🚀
Albert Cardona · @albertcardona
1668 followers · 1661 posts · Server mathstodon.xyz
@axoaxonic @lili Relatedly, this paper just got posted at the arxiv:
"Resurrecting Recurrent Neural Networks for Long Sequences" by Orvieto et al 2023 https://arxiv.org/abs/2303.06349
#neuroscience #machinelearning #rnn
Ulf Hamster · @ulf
52 followers · 15 posts · Server sigmoid.socialFeedback would be very appreciated
https://osf.io/bgkpv
#snn #sparseneuralnetwork #pretraining #recurrentneuralnetwork #rnn #dynamicsparsity
#snn #sparseneuralnetwork #pretraining #recurrentneuralnetwork #rnn #dynamicsparsity
Frank der Entspannende 🥾👟🇺🇦 · @DerEntspannende
364 followers · 2032 posts · Server digitalcourage.social
Eben wollte ich mich fast für mein erstes #Barcamp in diesem Jahr anmelden. Jetzt habe ich einen Artikel über meine Überlegungen geschrieben, warum ich vielleicht doch nicht teilnehme. Und ob ich nicht meine #Bahncard wegwerfen soll.
Das Barcamp? Ganz sicher ganz toll. An- und Abreise? Mit #Bus und #Bahn? Ähnlich sicher wie mit 10.000 Euro ins #Casino zu gehen.
Et voilà. 10. BarCamp Tourismusnetzwerk Rheinland-Pfalz – aber der ÖPNV!
https://derschreiben.de/10-barcamp-tourismusnetzwerk-rheinland-pfalz-aber-der-oepnv/
#rnn #trier #mainz #opnv #casino #bahn #bus #bahncard #barcamp
Machine Revolution · @machine_revolution
4 followers · 30 posts · Server masto.ai
https://github.com/BlinkDL/ChatRWKV Recurrent Neural Network that claims to match transformers like ChatGPT for text generation while using less resources.
Jascha Achterberg · @achterbrain
898 followers · 350 posts · Server mastodon.socialLooking forward to present my work on how spatially-embedded recurrent neural networks reveal widespread links between structural and functional neuroscience findings at the SNUFA Seminar Series:
This Wednesdays, 1st FEB, 15:00 CET
🧠🤖🌐
#neuroscience #ai #rnn #neuralnetwork
Published papers at TMLR · @tmlrpub
488 followers · 163 posts · Server sigmoid.social
Investigating Action Encodings in Recurrent Neural Networks in Reinforcement Learning
Matthew Kyle Schlegel, Volodymyr Tkachuk, Adam M White, Martha White
n)body · @n0body
26 followers · 138 posts · Server hachyderm.io
mikeiavelli · @mikeiavelli
9 followers · 26 posts · Server mathstodon.xyzInteresting approach:
Analyze trained networks by linearizing the dynamics around approximate fixed points.
"Reverse engineering recurrent networks for sentiment classification reveals line attractor dynamics", 2019
https://arxiv.org/abs/1906.10720
#rnn #nn #neuralnetworks #ml #machinelearning
RecursiveNeuron :verified: · @RecursiveNeuron
156 followers · 196 posts · Server techhub.social
Recurrent Neural Networks (RNN) for Language Modeling in Python course
#MachineLearning #RNN #pythonprogramming #python #NeuralNetworks
#machinelearning #rnn #pythonprogramming #python #neuralnetworks