The font of $$x + y$$ is wrong. In general, $$ ... $$ are unintuitive. generally, they would be expected to render some sort of latex equation (if they do anything). They do create a new line + center justify things, but don't render basic things.

inserting a link [[can I link to a page]] creates a link successfully, but it is to

alpha.weaver.sh/can%20I%20li...

which is likely never useful.

note that remarkable's products are not particularly good as ebook readers (their software support for it just really isn't there). I haven't heard of that changing with this new product.

For a quick proof, write \binom{N}{k} = \sum_{S\subseteq [n], |S| = k} (\prod_{i\in S} X_i) for indicators X_i, then it follows from linearity of expectation+independence.

E[\binom{N}{k}] = \sum_{S\subseteq [n], |S| = k}E[(\prod_{i\in S} X_i)] = \sum_{S\subseteq [n], |S| = k} p^k = \binom{n}{k}p^k

is your thought that they serve as an alternative avenue for "positive feedback" outside of work (e.g. "you made the number really big!!!" with a screen shake effect), or that they have a lower barrier to entry so are easier to engage with or something?

Worth clarifying those people are Australian, which makes the focus on them by right-wing Americans particularly weird.

there is a pokemon that is a possessed lawnmower

bulbapedia.bulbagarden.net/wiki/Rotom_(...

another pokemon is a castle wall tower (like a "Rook" in chess)

pokemondb.net/pokedex/stak...

the "PKE via paint mixing" is probably better. I've seen it various places, but for example

maths.straylight.co.uk/archives/108

Actually doing OTP/stream ciphers/RSA by hand seems mostly like an exercise in tedium though.

Toyota, as a company, is one of the biggest financial supporters of politicians stating the 2020 election was stolen.

www.motortrend.com/news/toyota-...

They have a bunch of unearned goodwill among libs due to the Prius specifically. Even for investments, their hydrogen fuel cell stuff was dumb af

this is billed as a way to support women in chess, via women-specific events. Chess is also notorious for having issues w/ sexual harassment of women though.

There have been some top women players who played in the "open" tournaments though and done very well, namely Judit Polgar.

It's especially funny because the idea of STEM revolutionizing farming *is* reasonable. It just happened over a century ago, and was a significant driver for the development of statistics as a discipline (well it, and eugenics).

What is the condition that rules out them agreeing to the constant yes/constant no strategy beforehand?

You can also use various web extensions to "force" you to use the subscription page only. I've been liking "Untrap for Youtube". Only subscription/channel webpages are available, can do things like turn off thumbnails, views/bunch of other information viewers don't care about, etc.

I've tested all of the following models locally

the 70b distill (you mentioned) is censored
the 32b distill is censored
the 14b distill is censored
the 8b distill is **not** censored (output attached)
the 1.5b and 7b distills are not censored, but doesn't appear to know what "tank man" means.

That isn't the model people mean when they talk about "Deepseek R1". Instead, it is the 671B MoE model. Querying "tank man" to it using a non-Chinese provider gives me something that seems reasonable

haven't tried it but saw this a few days ago

thomwiggers.nl/post/emoji-i...

is there a particular reason you appeal to Y being discrete (vs Markov's inequality) during the proof of fact 5?

KY sampling was explored some in lattice cryptography initially, but making it constant time hurts parameters fairly badly. You can read a description of it (and some other "generic" samplers) in section 2 of

eprint.iacr.org/2019/068.pdf

The bound linked (bound on the efficiency of Knuth-Yao sampling) requires either a certain auxiliary data structure to be built first (which only suffices for sampling a nearby distribution in TV dist) , or p(x,i), the ith binary digit of the pmf p(x), to be efficiently computable.