How big does a neural network need to be? | Oriol Vinyals and Lex Fridman

Показать описание

Please support this podcast by checking out our sponsors:

GUEST BIO:
Oriol Vinyals is the Research Director and Deep Learning Lead at DeepMind.

PODCAST INFO:

SOCIAL:

Рекомендации по теме

Комментарии

shouldn't it be how dynamic does a neural network need to be? Deterministically computing a whole bunch of weights seems to preclude employing higher order dynamic feedbacks. Feedbacks like having groups of parameters be able to auto bias groups of weights real time. A bit like the system has an introspective awareness that can gainfully generate complex algorithmic strategies above the weights, to say more efficiently make outputs conform to certain requirements rather than having screeds of training to fill out the solution space. Such that, the AI could now take inputs reimagine them according to some meta interpretation, then reinput a newly transformed meta state, process again to make further decisions based on some other meta criterions to internally bias the model, then maybe output, reinput, or perhaps self modify portions of the model so the AI is in a constantly evolving quest for some sense of meaning about the whole shebang lol...

blengi

Oriol Vinyals for some reasons looks like an absolute gigachad

polyloly

Stuart Hameroff would be a great guest, Orc OR is a compelling theory and raises the goal-post for number of connections needed to emulate human cognition.

deadlevelled

It's strange, we essentially have built the perfect billiards table, racked the balls and set the cue. The question becomes how do you make the game play itself with the only input being your break?

compositestechbb

How mathematics views neural, Till Now there are no straight study, a Basic and theorems put it,
It's not my field and my first look, I don't like a computer_Sci view, the neutral It's too much lack of pure mathematics,

__hannibaalbarca__

How big does a neural network need to be? | Oriol Vinyals and Lex Fridman

But what is a neural network? | Chapter 1, Deep learning

Neural Network In 5 Minutes | What Is A Neural Network? | How Neural Networks Work | Simplilearn

Neural Networks Explained in 5 minutes

Explained In A Minute: Neural Networks

The Essential Main Ideas of Neural Networks

How Many Hidden Layers and Neurons does a Neural Network Need

Neural Network Architectures & Deep Learning

I Built a Neural Network from Scratch

Day40: Basic Machine Learning-Weather Forecasting using Neural Network - Neural Network Model Design

Why Neural Networks can learn (almost) anything

What Do Neural Networks Really Learn? Exploring the Brain of an AI Model

What is a Neural Network?

Deep Learning | What is Deep Learning? | Deep Learning Tutorial For Beginners | 2023 | Simplilearn

Gradient descent, how neural networks learn | Chapter 2, Deep learning

Understanding Neural Networks and AI

Watching Neural Networks Learn

Liquid Neural Networks, A New Idea That Allows AI To Learn Even After Training

Attention for Neural Networks, Clearly Explained!!!

Neural Network Simply Explained | Deep Learning Tutorial 4 (Tensorflow2.0, Keras & Python)

Transformer Neural Networks, ChatGPT's foundation, Clearly Explained!!!

Why Do Tree Based-Models Outperform Neural Nets on Tabular Data?

What are Neural Networks || How AIs think

What is a Neural Network | Neural Networks Explained in 7 Minutes | Edureka

Recurrent Neural Networks (RNNs), Clearly Explained!!!