Parler Mark
Close Side Navigation
Parler Logo
Profile Cover Photo
Yannic Kilcher YannicKilcher I make Videos

馃敟New Video馃敟 FFT Magic馃獎Fourier Neural Operators speed up PDE solvers by orders(!) of magnitude 馃く Trained once, solve entire PDE families for any discretization!馃帀Watch to find out more鈴笍

#ai #science #news #technology #research #machinelearning #education #deeplearning #artificialintelligence #datascience #stem #ml #engineering #video #tech #youtube #paper #parler #technews

Fourier Neural Operator for Parametric Partial Differential Equations (Paper Explained) #ai #research #engineering Numerical solvers for Partial Differential Equations are notoriously slow. They need to evolve their state by tiny steps in order to stay accurate, and they need to repeat... https://youtu.be/IaS72aHrJKE

馃敟New Video馃敟 FFT Magic馃獎Fourier Neural Operators speed up PDE solvers by orders(!) of magnitude 馃く Trained once, solve entire PDE families for any discretization!馃帀Watch to find out more鈴笍

#ai #science #news #technology #research #machinelearning #education #deeplearning #artificialintelligence #datascience #stem #ml #engineering #video #tech #youtube #paper #parler #technews

Fourier Neural Operator for Parametric Partial Differential Equations (Paper Explained) #ai #research #engineering Numerical solvers for Partial Differential Equations are notoriously slow. They need to evolve their state by tiny steps in order to stay accurate, and they need to repeat... https://youtu.be/IaS72aHrJKE

Soccer AI FAILS and mixes up ball and referee's bald head. If only everyone had written more broader impact statements, this could have been avoided.

#ai #science #tech #research #news #artificialintelligence #machinelearning #ml #technologynews #parler #stem #technews #parleruse

[News] Soccer AI FAILS and mixes up ball and referee's bald head. This soccer camera is operated by an AI to track the ball. However, the AI has an interesting failure mode and repeatedly mixes up the ball with the bald head of a referee. This raises some... https://youtu.be/i_p5wLoCCiw

馃コVideo Time馃摻锔廇re all models created equal? Turns out no, not even if they perform equally well on the test(!) set馃く@GoogleAI Paper finds large variations among all "best" models & names underspecification as the problem!馃敟

#ai #research #news #tech #science #youtube #video #eduation #explainervideo #machinelearning #artificialintelligence #deeplearning #datascience #ml #stem #biology #technologynews #technology #innovation #education #data #google #optimization #genderbias #parler #parleruse #techtyrants

Underspecification Presents Challenges for Credibility in Modern Machine Learning (Paper Explained) #ai #research #machinelearning Deep Learning models are often overparameterized and have many degrees of freedom, which leads to many local minima that all perform equally well on the test set. But... https://youtu.be/gch94ttuy5s
Show Comments
Hide Comments

Is there a printed version?

馃帀 New Video 馃帀 Knowledge Graphs are very expensive to make, they need human experts. Or do they? 馃 What if we replaced them with BERT or GPT-2? 馃く Turns out, works really well, all without training! 馃コ

#ai #nlp #science #youtube #research #machinelearning #datascience #education #stem #news #paper # #deeplearning #artificialintelligence #ml #video #innovation #tech #data #technology #news #parler

Language Models are Open Knowledge Graphs (Paper Explained) #ai #research #nlp Knowledge Graphs are structured databases that capture real-world entities and their relations to each other. KGs are usually built by human experts, which costs considerable... https://youtu.be/NAJOZTNkhlI
Show Comments
Hide Comments

Interesting!

馃敟New Video馃敟 Linear Attention! Unbiased Estimator! Random Features! Orthogonal Features! Low Variance! Tight Bounds! Kernels! Backw. Compatible! The PERFORMER has it all馃く Watch!馃挭

#ai #research #science #technology #machinelearning #tech #deeplearning #artificialintelligence #news #datascience #ml #stem #biology #technologynews #deepmind #innovation #education #youtube #video #google #parler #technews

Rethinking Attention with Performers (Paper Explained) Transformers have huge memory and compute requirements because they construct an Attention matrix, which grows quadratically in the size of the input. The Reformer is a model that uses random... https://youtu.be/xJrKIPwVwGM

馃敟New Video馃敟 LambdaNetworks capture long-range interactions as linear functionals馃く Super complicated, basically Transformers without the giant memory requirements 馃コ New SotA on ImageNet! 馃挭 Watch Now!

#ICLR2021 #ai #research #tech #news #artificialintelligence #machinelearning #deeplearning #ml #science #datascience #technologynews #technology #video #youtube #education #parler

LambdaNetworks: Modeling long-range Interactions without Attention Transformer, having already captured NLP, have recently started to take over the field of Computer Vision. So far, the size of images as input has been challenging, as the Transformers' Attention... https://youtu.be/3qxJ2WD8p4w

Sunday Video馃 Choosing an Optimizer for Deep Learning is hard and often just done based on what other people do. Can we do better? This paper compares 14 of the most popular optimizers on 8 different DL problems in a 35K-run benchmark. 馃く

#ai #tech #technology #news #science #youtube #machinelearning #stem #education #artificialintelligence #deeplearning #optimization #parler

Descending through a Crowded Valley -- Benchmarking Deep Learning Optimizers (Paper Explained) #ai #research #optimization Deep Learning famously gives rise to very complex, non-linear optimization problems that cannot be solved analytically. Therefore, the choice of a suitable optimization... https://youtu.be/DiNzQP7kK-s

馃敟New Video馃敟Convolutions are DEAD as Transformers continue to ruin absolutely everything 馃槺 New SotA on ImageNet, VTAB, etc using only Transformers + massive data 馃憫 Also Peer Review is broken. Watch Now!馃憖

#ai #research #tech #machinelearning #news #education #technology #machinelearning #deeplearning #artificialintelligence #science #datascience #video #youtube #innovation #data #stem #parler #parlerusa #technologynews

An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (Paper Explained) #ai #research #transformers Transformers are Ruining Convolutions. This paper, under review at ICLR, shows that given enough data, a standard Transformer can outperform Convolutional Neural Networks... https://youtu.be/TrdevFK_am4

馃敟New Video馃敟Convolutions are DEAD as Transformers continue to ruin absolutely everything 馃槺 New SotA on ImageNet, VTAB, etc using only Transformers + massive data 馃憫 Also Peer Review is broken. Watch Now!馃憖

#ai #research #tech #machinelearning #news #education #technology #machinelearning #deeplearning #artificialintelligence #science #datascience #video #youtube #innovation #data #stem #parler #parlerusa #technologynews

An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (Paper Explained) #ai #research #transformers Transformers are Ruining Convolutions. This paper, under review at ICLR, shows that given enough data, a standard Transformer can outperform Convolutional Neural Networks... https://youtu.be/TrdevFK_am4

New Video 馃コ This paper uses lots of compute to learn a single, unified LSTM-based optimizer on over 6000(!) different tasks, then uses that optimizer to TRAIN ITSELF! We're in full meta-land 馃槺

#ai #research #machinelearning #news #technology #technologynews #tech #youtube #education #deeplearning #metalearning #optimization #google #science #artificialintelligence #biology #ml #stem #video #parler #parlerusa #parleruk #techtyrants

Training more effective learned optimizers, and using them to train themselves (Paper Explained) #ai #research #optimization Optimization is still the domain of hand-crafted, simple algorithms. An ML engineer not only has to pick a suitable one for their problem but also often do grid-search... https://youtu.be/3baFTP0uYOc

I love opinions馃憤This paper by @sarahookr has one: What if current big research ideas are big not because of their brilliance, but because they run well on GPUs?馃And how many brilliant ideas are never considered, just because they don't?馃Video here:

#ai #science #youtube #research #machinelearning #education #society #news #deeplearning #biology #artificialintelligence #stem #technology #tech #innovation #video #parler

The Hardware Lottery (Paper Explained) #ai #research #hardware We like to think that ideas in research succeed because of their merit, but this story is likely incomplete. The term "hardware lottery" describes the fact that certain... https://youtu.be/MQ89be_685o

I love opinions馃憤This paper by @sarahookr has one: What if current big research ideas are big not because of their brilliance, but because they run well on GPUs?馃And how many brilliant ideas are never considered, just because they don't?馃Video here:

#ai #science #youtube #research #machinelearning #education #society #news #deeplearning #biology #artificialintelligence #stem #technology #tech #innovation #video #parler

The Hardware Lottery (Paper Explained) #ai #research #hardware We like to think that ideas in research succeed because of their merit, but this story is likely incomplete. The term "hardware lottery" describes the fact that certain... https://youtu.be/MQ89be_685o

I love opinions馃憤This paper by @sarahookr has one: What if current big research ideas are big not because of their brilliance, but because they run well on GPUs?馃And how many brilliant ideas are never considered, just because they don't?馃Video here:

#ai #science #youtube #research #machinelearning #education #society #news #deeplearning #biology #artificialintelligence #stem #technology #tech #innovation #video #parler

The Hardware Lottery (Paper Explained) #ai #research #hardware We like to think that ideas in research succeed because of their merit, but this story is likely incomplete. The term "hardware lottery" describes the fact that certain... https://youtu.be/MQ89be_685o

Chess Time!鈾燂笍馃憫馃帬鉁濓笍馃椉What if you couldn't castle?馃 DeepMind investigates this & other rule changes using #AlphaZero to simulate millions of hours of human experience 馃敟 Watch the Video to find out more!

#ai #machinelearning #deeplearning #deepmind #youtube #education #tech #technology #science #news #alphago
#research #parler

Assessing Game Balance with AlphaZero: Exploring Alternative Rule Sets in Chess (Paper Explained) #ai #chess #alphazero Chess is a very old game and both its rules and theory have evolved over thousands of years in the collective effort of millions of humans. Therefore, it is almost impossible to... https://youtu.be/O1b0cbgpRBw

Video Time 馃コ Text Summarization is hard. In train, log-likelihood approaches are limited to reproduce given references. In eval, ROUGE fails to give accurate signals. This paper by @OpenAI uses human-in-the-loop and deep RL for both train and eval! 馃馃鈥嶐煢梆煂 #ai #tech #karma #parler #parlerusa #science #youtube #technology #machinelearning #deeplearning #programming #data #robots #openai

Learning to summarize from human feedback (Paper Explained) #summarization #gpt3 #openai Text Summarization is a hard task, both in training and evaluation. Training is usually done maximizing the log-likelihood of a human-generated reference summary, while... https://youtu.be/vLTmnaMpQCs

馃敟New Video馃敟 Self-Classifying MNIST: Using Neural Cellular Automata, each pixel can only talk to its neighbors, but together, they find out what class they are馃 Very cool! 馃挭馃槑#ai #tech #uvm #parler #news #parlerusa #youtube #education #machinelearning #science #technology

Self-classifying MNIST Digits (Paper Explained) #ai #biology #machinelearning Neural Cellular Automata are models for how living creatures can use local message passing to reach global consensus without a central authority. This paper teaches... https://youtu.be/EbHUU-gLyRA

New Video 馃コ Transformers are coming for Images 馃槺 Axial-DeepLab combine learned Positional Embeddings w/ Axial Attention and get SotA on Segmentation with a fully Attentional model! No Convolutions 馃悙#ai #tech #technology #youtube #parler #parlerusa #parleruk #parkerksa #fatkidclub #twexit #techtyrants #bitcoin #videogames #meme #news

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation (Paper Explained) Convolutional Neural Networks have dominated image processing for the last decade, but transformers are quickly replacing traditional models. This paper proposes a fully attentional model for images... https://youtu.be/hv3UO3G0Ofo

#news #parler #parlerusa #tech

#parler #news #tech

New Video 馃コ Transformers are coming for Images 馃槺 Axial-DeepLab combine learned Positional Embeddings w/ Axial Attention and get SotA on Segmentation with a fully Attentional model! No Convolutions 馃悙#ai #tech #technology #youtube #parler #parlerusa #parleruk #parkerksa #fatkidclub #twexit #techtyrants #bitcoin #videogames #meme #news

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation (Paper Explained) Convolutional Neural Networks have dominated image processing for the last decade, but transformers are quickly replacing traditional models. This paper proposes a fully attentional model for images... https://youtu.be/hv3UO3G0Ofo

#news #parler #parlerusa #tech

New Video 馃コ Transformers are coming for Images 馃槺 Axial-DeepLab combine learned Positional Embeddings w/ Axial Attention and get SotA on Segmentation with a fully Attentional model! No Convolutions 馃悙#ai #tech #technology #youtube #parler #parlerusa #parleruk #parkerksa #fatkidclub #twexit #techtyrants #bitcoin #videogames #meme #news

Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation (Paper Explained) Convolutional Neural Networks have dominated image processing for the last decade, but transformers are quickly replacing traditional models. This paper proposes a fully attentional model for images... https://youtu.be/hv3UO3G0Ofo