Machine Learning

1 readers

1 users here now

Community Rules:

Be nice. No offensive behavior, insults or attacks: we encourage a diverse community in which members feel safe and have a voice.
Make your post clear and comprehensive: posts that lack insight or effort will be removed. (ex: questions which are easily googled)
Beginner or career related questions go elsewhere. This community is focused in discussion of research and new projects that advance the state-of-the-art.
Limit self-promotion. Comments and posts should be first and foremost about topics of interest to ML observers and practitioners. Limited self-promotion is tolerated, but the sub is not here as merely a source for free advertisement. Such posts will be removed at the discretion of the mods.

founded 1 year ago

MODERATORS

communick@academy.garden

[P] JustANeuron - Understand Backpropagation Through an Interactive Visualization of a Single Neuron (alien.top)

submitted 11 months ago by xjustwaitx@alien.top to c/machinelearning@academy.garden

3 comments fedilink hide all child comments

I recently saw Karpathy's video on backpropagation, and at some point around the 40-minute mark he shows how to visualize backpropagation flowing through the computational graph of a single neuron, which I felt really made me understand automatic backpropagation intuitively for the first time. But I felt that I had to mess around with it for it to really sink in. So me and a friend quickly made an interactive, online version of the visualization, please tell me what you think!

https://amitalevy.com/justaneuron/

top 3 comments

sorted by: hot top controversial new old

[–] Infamous-Bank-7739@alien.top 1 points 11 months ago

Why do the multiplications and sums have a grad? Is this related to how autograd works? I understand it in equation terms, but don't see the necessity in computational terms

[–] xaist@alien.top 1 points 11 months ago

Thanks.

[–] theturnerk@alien.top 1 points 11 months ago

wow, this is fantastic! I've been trying to wrap my head around backpropagation for a while, and this visualization really helps make it click. Great work on creating this interactive version - it's super helpful for getting a hands-on understanding of the concept. Thanks for sharing!