HN user

britcruise

137 karma

artoftheproblem

Posts35
Comments14
View on HN
www.youtube.com 2y ago

The "Magic" of Bitcoin [video]

britcruise
1pts0
www.storyxperiential.com 3y ago

Story Xperiential with Pixar is a program for anyone to level up their skills

britcruise
2pts0
www.storyxperiential.com 4y ago

Pixar opens it's doors with Story Xperiential

britcruise
3pts1
www.youtube.com 6y ago

Manifold Hypothesis explained with paper folding analogy

britcruise
1pts0
www.youtube.com 6y ago

Why deep neural networks work (explained with paint and pianos)

britcruise
6pts0
www.youtube.com 6y ago

IEEE and Art of the Problem Explain Adi Shamir's Secret Sharing Paper from 1979

britcruise
1pts0
www.youtube.com 6y ago

Secret Sharing Explain Visually

britcruise
1pts0
docs.google.com 6y ago

Please find mistakes in this rough outline for a video on neural network history

britcruise
1pts0
news.ycombinator.com 6y ago

I have question about information theoretic view of backprop

britcruise
2pts2
www.youtube.com 6y ago

I spend a year on this video on Machine Learning and it's connection to nature

britcruise
1pts0
www.youtube.com 6y ago

New 3 hour series on how Theme Parks are engineered: “Imagineering in a Box”

britcruise
1pts0
www.youtube.com 7y ago

The Nature of Learning (Art of the Problem on Machine Learning)

britcruise
3pts0
www.youtube.com 7y ago

Key insights behind Deep Learning (distributed representations)

britcruise
1pts0
www.youtube.com 7y ago

Art of the Problem explains the key insight behind deep learning

britcruise
3pts0
www.youtube.com 7y ago

The Pattern Machine (Deep Learning Video Series by Art of the Problem)

britcruise
6pts0
www.youtube.com 7y ago

Art of the Problem announced a 7 part series on AI (teaser)

britcruise
2pts0
www.youtube.com 7y ago

Art of the Problem and IEEE are making videos on the most cited papers

britcruise
2pts0
www.youtube.com 7y ago

The beauty of lempel-ziv compression

britcruise
1pts0
www.youtube.com 7y ago

The beauty of lempel-ziv compression

britcruise
2pts0
www.youtube.com 7y ago

Art of the Problem posted a video on modern error correction (LDPC codes)

britcruise
3pts0
www.youtube.com 8y ago

I spent 16 months on this Bitcoin explainer. by Art of the Problem

britcruise
28pts9
www.youtube.com 8y ago

Over a year in the making, a 30m video on Bitcoin by Art of the Problem -TRAILER

britcruise
2pts0
www.youtube.com 8y ago

How Space-Time Codes Work (video by Art of the Problem and IEEE)

britcruise
1pts0
www.youtube.com 8y ago

Art of the Problem and IEEE posted new video series on modern Information Theory

britcruise
1pts0
www.youtube.com 8y ago

I just posted a new Art of the Problem video on P vs. NP and would love feedback

britcruise
1pts0
www.youtube.com 9y ago

Art of the Problem posted a new video on Turing Machines

britcruise
1pts0
www.youtube.com 9y ago

What is a Computer? (art of the problem)

britcruise
2pts0
www.youtube.com 9y ago

Abstraction, Logic and Aristotle – Video by Art of the Problem

britcruise
1pts0
www.youtube.com 9y ago

Art of the Problem Tackles LOGIC and ABSTRACTION

britcruise
1pts0
www.khanacademy.org 10y ago

We just published a new Pixar in a Box lesson on particle simulations

britcruise
19pts0

thanks for the comment.

Yes the performance boost in training is critical.

So what I'm saying is this performance boost is thanks to the switch to non-binary neuron (which allows a reversible operation to transmit a magnitude) - that's MOST important.

And separately ReLu are just better at this, because they are linear they don't have "vanishing edges" (which prevents the vanish/explode)

Separately I'm glad you brought up CNNs because CNN's are old, and go back to rosenblatt (1958), his perception had a first layer of local connections in it based on the findings in biological systems.

and of course that's because nature has found it's more efficient, and so the efficiency is huge.

but the point is CNN = fewer knobs to train.

and there are lots of simple ways to help reduce knobs - fix some connections - drop some connections

These are great points. You'll notice that video is actually a "bonus" step which isn't required to complete the final exercise in that lesson. We are still experimenting with how to go about "bonus" steps. In other lessons you'll see they are done in a multi-step article style. Such as in Animation: https://www.khanacademy.org/partner-content/pixar/animate/pa...

We also do this for the bonus step in Character Modeling (my favorite lesson!)

Our plan is to continue collecting more feedback on these different styles and find out what works best.

Hey HN! My name is Brit Cruise and I'm the lead content developer on the Khan Academy side of this equation (working alongside Tony DeRose from Pixar...plus a village of others), if you have any specific questions about this project I'll be happy to answer as best I can.