#tsetlinmachine search results

The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

LiteralLabs's tweet image. The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

Just achieved 91.82% peak accuracy on #FashionMNIST with a tiny #TsetlinMachine model using only 20 clauses per class (10 positive + 10 negative) through a modified algorithm and simple convolution. In 2019, CTM achieved 91.5% accuracy with 400 times more clauses — 8000 per class

ArtemHnilov's tweet image. Just achieved 91.82% peak accuracy on #FashionMNIST with a tiny #TsetlinMachine model using only 20 clauses per class (10 positive + 10 negative) through a modified algorithm and simple convolution. In 2019, CTM achieved 91.5% accuracy with 400 times more clauses — 8000 per class

Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

ArtemHnilov's tweet image. Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

Last night, I achieved a peak accuracy of 93.17% on the #FashionMNIST dataset using a tiny 20-clause per class #TsetlinMachine with a modified algorithm and convolution.

ArtemHnilov's tweet image. Last night, I achieved a peak accuracy of 93.17% on the #FashionMNIST dataset using a tiny 20-clause per class #TsetlinMachine with a modified algorithm and convolution.

Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

ArtemHnilov's tweet image. Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

ArtemHnilov's tweet image. The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

Our latest publication: Voice recognition with the #TsetlinMachine Video Demo 📽& Full-text 📜: tinyurl.com/KWSTMDEMO

That_JieLei's tweet image. Our latest publication:
Voice recognition with the #TsetlinMachine 

Video Demo 📽& Full-text 📜: tinyurl.com/KWSTMDEMO

Well done @That_JieLei for getting your article "From Arithmetic to Logic Based AI: a Comparative Analysis of Neural Networks and Tsetlin Machine" accepted @ICECS2020! Great insights on #ML #TsetlinMachine and #NN. @AdrianWheeldon @alexyakovlevncl @olegranmo.

RishadShafik's tweet image. Well done @That_JieLei for getting your article "From Arithmetic to Logic Based AI: a Comparative Analysis of Neural Networks and Tsetlin Machine" accepted @ICECS2020! Great insights on #ML #TsetlinMachine and #NN. @AdrianWheeldon @alexyakovlevncl @olegranmo.

Fantastic collaboration! @abu_bakar___'s & @TousifSRahman's internship work (Logic-based Intelligence for Batteryless Sensors) at Nokia Bell Labs is accepted for HotMobile'22. The paper shows methods of low-complexity #TsetlinMachine. Well done everyone! tinyurl.com/5bkefm47.

RishadShafik's tweet image. Fantastic collaboration! @abu_bakar___'s & @TousifSRahman's internship work (Logic-based Intelligence for Batteryless Sensors) at Nokia Bell Labs is accepted for HotMobile'22. The paper shows methods of low-complexity #TsetlinMachine. Well done everyone! tinyurl.com/5bkefm47.

Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS! I believe this might be a world record. Can anyone beat this result?

ArtemHnilov's tweet image. Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS!
I believe this might be a world record. Can anyone beat this result?

AURORA introduces an automated, CMOS compatible 8T SRAM architecture that performs wired-OR logic operations in-memory. Demonstrated using the #TsetlinMachine as a case study, the design achieves >100× higher throughput and >770× lower energy compared to existing methods.


Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

ArtemHnilov's tweet image. Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

ArtemHnilov's tweet image. I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.

RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.
RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.
RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.

After adjusting the booleanization threshold, I achieved a peak accuracy of 92.67% on the #FashionMNIST dataset using a tiny #TsetlinMachine model with 20 clauses per class and a simple #convolution.

ArtemHnilov's tweet image. After adjusting the booleanization threshold, I achieved a peak accuracy of 92.67% on the #FashionMNIST dataset using a tiny #TsetlinMachine model with 20 clauses per class and a simple #convolution.

An interview with @RishadShafik, @IconLrf supervisor @UniofNewcastle, has been published in Electronics Weekly. He discussed their recent paper on #TsetlinMachine #MachineLearning #Hardware to be published in Royal Society Philosophical Trans A : ow.ly/1BAj50AJT3o

IconLrf's tweet image. An interview with @RishadShafik, @IconLrf supervisor @UniofNewcastle, has been published in Electronics Weekly.   He discussed their recent paper on #TsetlinMachine #MachineLearning #Hardware to be published in Royal Society Philosophical Trans A : ow.ly/1BAj50AJT3o

Two CAIR papers at IEA-AIE 2020: * A Regression Tsetlin Machine with Integer Weighted Clauses for Compact Pattern Representation * Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing, Congratulations, Darshana and Saeed! #TsetlinMachine @uiagder

CairEnglish's tweet image. Two CAIR papers at IEA-AIE 2020:
* A Regression Tsetlin Machine with Integer Weighted Clauses for Compact Pattern Representation
* Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing, 
Congratulations, Darshana and Saeed!
#TsetlinMachine @uiagder

First Tsetlin Machine chip - mignonAI - gone into production. Congratulations to our friends at microSystems, Newcastle University! #TsetlinMachine #mignonAI #PervasiveAI #InterpretableAI @uiagder

CairEnglish's tweet image. First Tsetlin Machine chip - mignonAI - gone into production. Congratulations to our friends at microSystems, Newcastle University! #TsetlinMachine #mignonAI #PervasiveAI #InterpretableAI @uiagder

A thread for celebrating paper successes of our @nclmicrosystems #TsetlinMachine Working Group researchers who submitted to International Symposium on the Tsetlin Machine (ISTM) 2022. Well done, everyone! 🙌🙌🙌


The #TsetlinMachine coming to Rome in 2025

Brace yourselves for the fourth edition of International Symposium on the Tsetlin Machine (ISTM) in Rome in 2025. Until then it’s a goodbye from Pittsburgh.

RishadShafik's tweet image. Brace yourselves for the fourth edition of International Symposium on the Tsetlin Machine (ISTM) in Rome in 2025. Until then it’s a goodbye from Pittsburgh.


AURORA introduces an automated, CMOS compatible 8T SRAM architecture that performs wired-OR logic operations in-memory. Demonstrated using the #TsetlinMachine as a case study, the design achieves >100× higher throughput and >770× lower energy compared to existing methods.


The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

LiteralLabs's tweet image. The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

The #TsetlinMachine reached complex neural networks like Inception-v3, hitting 94.47% test accuracy (peaking at 94.63%) on #FashionMNIST! #MachineLearning


Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

ArtemHnilov's tweet image. Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

🚀 Building Tactyl – The New Open-Core Detection Engine 🔧 Powered by symbolic logic. Built for defenders I’m building Tactyl, a high-speed detection engine in Go that uses symbolic logic instead of black-box models. #CyberSecurity #TsetlinMachine #BlueTeam #SOC #ExplainableAI


Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS! I believe this might be a world record. Can anyone beat this result?

ArtemHnilov's tweet image. Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS!
I believe this might be a world record. Can anyone beat this result?

The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

ArtemHnilov's tweet image. The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

ArtemHnilov's tweet image. Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

Congratulations @OmarGGhazal for getting 'IMPACT : In-Memory ComPuting Architecture Based on Y-FlAsh Technology for Coalesced Tsetlin Machine Inference' accepted in the RS Phil. Trans. Our first-ever Y-flash based 180-nm #TsetlinMachine design. Preprint: arxiv.org/abs/2412.05327


Exciting news! The Tsetlin Machine is revolutionizing AI by reducing energy consumption, making it more sustainable for the environment. 🌱♻️ Let's embrace technology that serves humanity and the planet! #TsetlinMachine #AIinnovation unite.ai/a-game-changer…


Why are we passionate about #TsetlinMachine #AI? Reduced energy use (amongst other things). @UniteAi takes a deep dive into TMs and sustainable AI unite.ai/a-game-changer…


Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

ArtemHnilov's tweet image. Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

ArtemHnilov's tweet image. I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

The #TsetlinMachine coming to Rome in 2025

Brace yourselves for the fourth edition of International Symposium on the Tsetlin Machine (ISTM) in Rome in 2025. Until then it’s a goodbye from Pittsburgh.

RishadShafik's tweet image. Brace yourselves for the fourth edition of International Symposium on the Tsetlin Machine (ISTM) in Rome in 2025. Until then it’s a goodbye from Pittsburgh.


For fun, I trained a modified #TsetlinMachine with just 2 (Two!) clauses per class (1 positive and 1 negative) using convolution and achieved a peak accuracy of 91.75% on the #FashionMNIST dataset.


No results for "#tsetlinmachine"

The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

LiteralLabs's tweet image. The full program for Rome's International Symposium on the #TsetlinMachine (ISTM) is now out with no fewer than 3 Literal Labs staff in attendance

Our latest publication: Voice recognition with the #TsetlinMachine Video Demo 📽& Full-text 📜: tinyurl.com/KWSTMDEMO

That_JieLei's tweet image. Our latest publication:
Voice recognition with the #TsetlinMachine 

Video Demo 📽& Full-text 📜: tinyurl.com/KWSTMDEMO

Last night, I achieved a peak accuracy of 93.17% on the #FashionMNIST dataset using a tiny 20-clause per class #TsetlinMachine with a modified algorithm and convolution.

ArtemHnilov's tweet image. Last night, I achieved a peak accuracy of 93.17% on the #FashionMNIST dataset using a tiny 20-clause per class #TsetlinMachine with a modified algorithm and convolution.

Just achieved 91.82% peak accuracy on #FashionMNIST with a tiny #TsetlinMachine model using only 20 clauses per class (10 positive + 10 negative) through a modified algorithm and simple convolution. In 2019, CTM achieved 91.5% accuracy with 400 times more clauses — 8000 per class

ArtemHnilov's tweet image. Just achieved 91.82% peak accuracy on #FashionMNIST with a tiny #TsetlinMachine model using only 20 clauses per class (10 positive + 10 negative) through a modified algorithm and simple convolution. In 2019, CTM achieved 91.5% accuracy with 400 times more clauses — 8000 per class

Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

ArtemHnilov's tweet image. Achieved 101 million #MNIST predictions per second (throughput: 55.5 GB/s) with 98.10% accuracy on a CPU using the latest optimizations in the #TsetlinMachine library, Tsetlin.jl.

Well done @That_JieLei for getting your article "From Arithmetic to Logic Based AI: a Comparative Analysis of Neural Networks and Tsetlin Machine" accepted @ICECS2020! Great insights on #ML #TsetlinMachine and #NN. @AdrianWheeldon @alexyakovlevncl @olegranmo.

RishadShafik's tweet image. Well done @That_JieLei for getting your article "From Arithmetic to Logic Based AI: a Comparative Analysis of Neural Networks and Tsetlin Machine" accepted @ICECS2020! Great insights on #ML #TsetlinMachine and #NN. @AdrianWheeldon @alexyakovlevncl @olegranmo.

An interview with @RishadShafik, @IconLrf supervisor @UniofNewcastle, has been published in Electronics Weekly. He discussed their recent paper on #TsetlinMachine #MachineLearning #Hardware to be published in Royal Society Philosophical Trans A : ow.ly/1BAj50AJT3o

IconLrf's tweet image. An interview with @RishadShafik, @IconLrf supervisor @UniofNewcastle, has been published in Electronics Weekly.   He discussed their recent paper on #TsetlinMachine #MachineLearning #Hardware to be published in Royal Society Philosophical Trans A : ow.ly/1BAj50AJT3o

Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.

RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.
RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.
RishadShafik's tweet image. Check out our #HW #TsetlinMachine work "Learning automata based energy-efficient AI hardware design ...."; #OpenAccess in @RSocPublishing Phil. Trans. of The Royal Society A: tinyurl.com/y6oen29q. @AdrianWheeldon @That_JieLei #TousifRahman @alexyakovlevncl @olegranmo.

Highlight from recent #mignonAI #TsetlinMachine research: systematic prodigality (more states or clauses) provides a high-level of fault masking/mitigation capability w/o any additional design effort. Full paper on its way; meanwhile read: tinyurl.com/y74c8epq @nclmicrosystems

RishadShafik's tweet image. Highlight from recent #mignonAI #TsetlinMachine research: systematic prodigality (more states or clauses) provides a high-level of fault masking/mitigation capability w/o any additional design effort. Full paper on its way; meanwhile read: tinyurl.com/y74c8epq @nclmicrosystems
RishadShafik's tweet image. Highlight from recent #mignonAI #TsetlinMachine research: systematic prodigality (more states or clauses) provides a high-level of fault masking/mitigation capability w/o any additional design effort. Full paper on its way; meanwhile read: tinyurl.com/y74c8epq @nclmicrosystems

Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

ArtemHnilov's tweet image. Sentiment analysis on the #IMDB dataset achieved a test accuracy of 91.57% using the #TsetlinMachine.

I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

ArtemHnilov's tweet image. I achieved 73 million MNIST predictions per second on the #AMD #Ryzen 7950X3D CPU, with a 98.10% test accuracy, using the latest changes in the #TsetlinMachine library Tsetlin.jl.

The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

ArtemHnilov's tweet image. The Nash equilibrium was reached after 6,743 epochs of training the #TsetlinMachine on the #IMDB dataset. Accuracy stabilized at 90.34%, and none of the included literals changed further. Additionally, the training speed per epoch was fast.

Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

ArtemHnilov's tweet image. Achieved 94.02% peak test accuracy on #FashionMNIST using #TsetlinMachine with convolutional preprocessing! #MachineLearning #AI

Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS! I believe this might be a world record. Can anyone beat this result?

ArtemHnilov's tweet image. Just achieved 616 million #MNIST predictions per second on 240 threads using #TsetlinMachine — huge thanks to Joel from @NumascaleAS!
I believe this might be a world record. Can anyone beat this result?

After adjusting the booleanization threshold, I achieved a peak accuracy of 92.67% on the #FashionMNIST dataset using a tiny #TsetlinMachine model with 20 clauses per class and a simple #convolution.

ArtemHnilov's tweet image. After adjusting the booleanization threshold, I achieved a peak accuracy of 92.67% on the #FashionMNIST dataset using a tiny #TsetlinMachine model with 20 clauses per class and a simple #convolution.

Fantastic collaboration! @abu_bakar___'s & @TousifSRahman's internship work (Logic-based Intelligence for Batteryless Sensors) at Nokia Bell Labs is accepted for HotMobile'22. The paper shows methods of low-complexity #TsetlinMachine. Well done everyone! tinyurl.com/5bkefm47.

RishadShafik's tweet image. Fantastic collaboration! @abu_bakar___'s & @TousifSRahman's internship work (Logic-based Intelligence for Batteryless Sensors) at Nokia Bell Labs is accepted for HotMobile'22. The paper shows methods of low-complexity #TsetlinMachine. Well done everyone! tinyurl.com/5bkefm47.

First Tsetlin Machine chip - mignonAI - gone into production. Congratulations to our friends at microSystems, Newcastle University! #TsetlinMachine #mignonAI #PervasiveAI #InterpretableAI @uiagder

CairEnglish's tweet image. First Tsetlin Machine chip - mignonAI - gone into production. Congratulations to our friends at microSystems, Newcastle University! #TsetlinMachine #mignonAI #PervasiveAI #InterpretableAI @uiagder

Two CAIR papers at IEA-AIE 2020: * A Regression Tsetlin Machine with Integer Weighted Clauses for Compact Pattern Representation * Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing, Congratulations, Darshana and Saeed! #TsetlinMachine @uiagder

CairEnglish's tweet image. Two CAIR papers at IEA-AIE 2020:
* A Regression Tsetlin Machine with Integer Weighted Clauses for Compact Pattern Representation
* Increasing the Inference and Learning Speed of Tsetlin Machines with Clause Indexing, 
Congratulations, Darshana and Saeed!
#TsetlinMachine @uiagder

Loading...

Something went wrong.


Something went wrong.


United States Trends