naz

self studying math:
done: hammack, spivak, lay & lay
now: colley
next: simmons, blitzstein, abbott, wasserman

[naz @AlgorithmicBot]

Implication, necessary and sufficient conditions are usually taught inadequately

Here is how I would teach these concepts: https://t.co/GYhWVy8Jtd

[naz @AlgorithmicBot]

colley's vector calc notation goes crazy

let's define function f, whose co-ordinate inputs are x, y, z; let's also define function x (just like a co-ordinate, but boldfaced), and let's have inputs s, t

yikes

[naz @AlgorithmicBot]

problem from Putnam, that if you have the right idea can be solved pretty quickly:

Let A and B be different n by n matrices with real entries. If A^3 = B^3 and A^2 B = B^2 A, can A^2 + B^2 be invertible?

[naz @AlgorithmicBot]

@bongrandp @HeMuyu0327 obviously. i meant even the formatting differs, so it looks like it's something completely different

[Gwart @GwartyGwart]

Someone should build a version of polymarket where you can bet on something and then you win the bet if that thing happens

[naz @AlgorithmicBot]

@norbertbodziony @cobie what group was this deal sourced through, don't think I am part of it. down bad.

and congrats!😀

[naz @AlgorithmicBot]

tbh, i haven't tried anki as extensively, so wouldn't be able to compare properly. but i really love mochi's ui. i know that it's easier to have different categories of cards in anki, like you can create huge reference cards, bunch of example cards and not have them in active recall; whereas with mochi you have to use the so called "tags" to distinguish between different types of cards, but they will still be part of active recall. also, anki is supposedly better for when you have tons and tons of cards. but then, if you do have so many cards, then you might not be able to have them as part of the active recall due to the sheer number (like you won't be able to do 200-300 cards daily).

in any case, hopefully mochi fixes their card deck sharing and i will share mine here / and in the blog post

[naz @AlgorithmicBot]

linear algebra by lay and lay done

took me just under 6 weeks, produced around ~1000 flashcards

will write a blog post about how I self study math soon

next on: hubbard's vector calclus. this one is going to likely take longer. but seems first 4 chapters overlap a ton with linear algebra anyway

[naz @AlgorithmicBot]

@00pium_math whenever I try to make them public, heroku app that hosts them (I use @MochiCardsApp for flashcards) just crashes. I don't think they were anticipating someone to have so many cards in a single deck

[naz @AlgorithmicBot]

@ThePumponomics time to sell everything and just live off the yield😄

[naz @AlgorithmicBot]

@Memetic_Theory very strange paper

[✧ 𝕀𝔸𝕄𝔸𝕀 ✧ @iamai_omni]

Jane Street 的这篇文章写得也太漂亮了，把抽象群论和实际 ML 模型结合得非常漂亮，读完会有“原来位置编码的空间这么受限”的感觉。强烈推荐给对数学 + 深度学习感兴趣的人。

https://t.co/ATEwVOVcJj

[naz @AlgorithmicBot]

interference disappears the moment we ask which path it took? time to join the instrumentalists

[naz @AlgorithmicBot]

took me an hour to think through a => b, but now when I read it, seems "trivial" 🥲

[arya @AJakkli]

I really liked Eric's take on why alpha go is profound: A 10-layer network can only do 10 sequential steps of thinking, by construction. And yet those 10 steps can "amortize and approximate to very high fidelity a nearly intractable search problem."

[naz @AlgorithmicBot]

@fchollet going to be starting hubbard's vector calculus soon, will be epic

[pmxt @devpmxt]

All archive data now live via API

[naz @AlgorithmicBot]

@cryptopatrick @davbelian

[naz @AlgorithmicBot]

@banteg 😂

[PsikoBilim @Psikobilim_]

Bir kişinin zekâ, çalışkanlık ve duygusal olarak dengeli olma özelliklerinin çok üst düzeyine aynı anda sahip olma olasılığı milyonda 85'tir.