Live data from Hacker News

Deriving the SVD (Single Value Decomposition) from scratch

stillthinking.net

11–19 of 19 posts

Re: Deriving the SVD (Single Value Decomposition) from scratch

#11
post #7

This is a decent derivation, if a little verbose, but I find myself doubting the disclaimer at the top: > Disclaimer: no AI was used to write this. Any errors, awkward sentences, and weird tangents are 100% organic, free-range, and human-made. If you're writing by hand and trying to avoid accusations of AI-generated content, it's really easy not to write bulleted lists where each item starts with a boldface summary l…

> Same transformation. Same map on every actual vector. But now it looks like an arbitrary tangle of scaling, shearing, and sign flips — you’d never guess it was just “stretch x by 3.” I have a very hard time imagining a human writing this if they weren't intentionally attempting to sound like an LLM.

And,

> This is the engine behind image compression

SVD+truncate (or any other low rank approximation) is absolutely not used in any real image compression. Of course, it is a popular classroom project to demonstrate SVD. It's O(n^3) ffs.

Only AI would call SVD the "engine behind image compression", lol.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#12
These day's I'm super-into information theory and entropy, so I liked the connection made at the end. I'm a visual person, and I found this schematic of SVD on the wikipedia page very insightful [1].

[1]: https://en.wikipedia.org/wiki/Singular_value_decomposition#/...

Re: Deriving the SVD (Single Value Decomposition) from scratch

#13

This is a decent derivation, if a little verbose, but I find myself doubting the disclaimer at the top: > Disclaimer: no AI was used to write this. Any errors, awkward sentences, and weird tangents are 100% organic, free-range, and human-made. If you're writing by hand and trying to avoid accusations of AI-generated content, it's really easy not to write bulleted lists where each item starts with a boldface summary l…

Im not trying anything. I just put as a joke. Im using AI to learn, but I wrote the structure and the references, and remodeled some parts with AI, yes, but I wrote the base input text by hand, in the sense AI would not be needed, but it helps me as English is not my first language.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#14

This is a decent derivation, if a little verbose, but I find myself doubting the disclaimer at the top: > Disclaimer: no AI was used to write this. Any errors, awkward sentences, and weird tangents are 100% organic, free-range, and human-made. If you're writing by hand and trying to avoid accusations of AI-generated content, it's really easy not to write bulleted lists where each item starts with a boldface summary l…

So many shibboleths. Zero chance it's not AI.

Except anyone can use bullets.

And except the section title is basically me resuming what was seeing before:

Taking stock: what we managed to extract

So you want me to use what? Give me a break.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#15
post #7

Earlier quoted context omitted.

> Same transformation. Same map on every actual vector. But now it looks like an arbitrary tangle of scaling, shearing, and sign flips — you’d never guess it was just “stretch x by 3.” I have a very hard time imagining a human writing this if they weren't intentionally attempting to sound like an LLM.

And, > This is the engine behind image compression SVD+truncate (or any other low rank approximation) is absolutely not used in any real image compression. Of course, it is a popular classroom project to demonstrate SVD. It's O(n^3) ffs. Only AI would call SVD the "engine behind image compression", lol.

Im fully aware. But since you came here to criticize only, thanks.

Its one of the simple forms of lossy compression, so you could have just point out that constructively, but hey this is internet.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#16
post #6

This is a decent derivation, if a little verbose, but I find myself doubting the disclaimer at the top: > Disclaimer: no AI was used to write this. Any errors, awkward sentences, and weird tangents are 100% organic, free-range, and human-made. If you're writing by hand and trying to avoid accusations of AI-generated content, it's really easy not to write bulleted lists where each item starts with a boldface summary l…

Yeah it's 100% bullshit. The biggest red flag here isn't the use of AI, it's the author thinking we're dumb enough not to be able to detect it easily. That itself is dumb and makes this whole article easy to disregard.

Let me give you an idea, just dont read it then. Im not monetizing anything btw.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#17
post #16
post #6

Earlier quoted context omitted.

Yeah it's 100% bullshit. The biggest red flag here isn't the use of AI, it's the author thinking we're dumb enough not to be able to detect it easily. That itself is dumb and makes this whole article easy to disregard.

Let me give you an idea, just dont read it then. Im not monetizing anything btw.

No. Worse. You're wasting our time

Re: Deriving the SVD (Single Value Decomposition) from scratch

#18
post #15

Earlier quoted context omitted.

And, > This is the engine behind image compression SVD+truncate (or any other low rank approximation) is absolutely not used in any real image compression. Of course, it is a popular classroom project to demonstrate SVD. It's O(n^3) ffs. Only AI would call SVD the "engine behind image compression", lol.

Im fully aware. But since you came here to criticize only, thanks. Its one of the simple forms of lossy compression, so you could have just point out that constructively, but hey this is internet.

It would have been in a different tone had it been human written. In fact, if it had not been written with AI, there is no chance SVD gets called the fancy "engine behind image compression", the blog would have said exactly what you said in the comment above: "Its one of the simple forms of lossy compression, here's a simple before/after on a sample image" or whatever.

Re: Deriving the SVD (Single Value Decomposition) from scratch

#19
post #16

Earlier quoted context omitted.

Let me give you an idea, just dont read it then. Im not monetizing anything btw.

No. Worse. You're wasting our time

Lol, its funny how people think they know about others peoples lives... and speaking of wasting time...
Post reply on HN