naz

Implication, necessary and sufficient conditions are usually taught inadequately Here is how I would teach these concepts: blog.naz.ooo

@PessoaBrain dynamix looks pretty remarkable

𝗥𝗲𝗰𝗼𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗶𝗻𝗴 𝗱𝘆𝗻𝗮𝗺𝗶𝗰𝗮𝗹 𝘀𝘆𝘀𝘁𝗲𝗺𝘀 Looking forward to diving into this one. By Durstewitz et al. arxiv.org/abs/2602.16864

@banteg this only applies to threshold cryptography?

i hope my post is wrong because otherwise we are at the cryptography breaking stage, which is terrifying.

@chris_juravich finishing up Lay's Linear Algebra right now myself, but this is on my reading list as well. Sheldon Axler also has a supporting material (quite concise) for real analysis required for this book

Analysis posts incoming. If you don’t already have Prof Axler’s book, you should get it if you’re at all interested in math.

Confession: I never had a single work-related sleepless night or ever pulled an all-nighter during my career incl. PhD. Don’t sacrifice your health. Sleep is a superpower — your brain on 8hrs of sleep is a lot smarter than your brain on sleep deprivation. Don’t listen to people who tell you to chronically sacrifice sleep for work. Sacrificing sleep for your kids/family is a different story.

me: claude, read every markdown file in the repo claude: i have read every markdown file in the repo honest caveat: i haven't read all of x.md, i am comfortable with the gap me: 🤦‍♂️

@SalsaTekila OR it's a "bullish market" because of all of those listed things (except, for quantum)

The fact that we’re in a bullish market and not dumping despite the 10,000 reasons we’ve had; from Israel attacking Iran, the war escalating, oil, quantum, Saylor literally scaring every potential new investor away while hinting they might sell some coins… Scarily bullish.

@atmoio all of a sudden my claude started using this in every single response, it was driving me nuts

why is everyone saying load-bearing all of a sudden

@SalsaTekila i also make it make milestone plan, so that even if i compact it, it will know where we are at in terms of progress: what was done, and what needs to be done. and i make it plan every milestone before it just blindly implements

When I build stuff with Claude, 5% of the work is the idea, 5% is writing the prompt, and 90% is reading and answering questions, because I always ask: "Ask me questions until you are 100% sure what I want, do not make guesses." Am I doing it right?

@calvinfroedge things people likely say in a bubble be fearful when people are greedy and load up with puts

The Nasdaq will never crash Any day it is up less than 1% is an opportunity to buy the dip We are in a new paradigm AI is a Cambrian explosion of wealth and productivity All previous economic regimes and measuring systems are now irrelevant and outdated

Saylor is going to kill BTC because he will own too many bitcoins but also he's going to kill BTC because he will sell some bitcoins

"our results suggest that understanding neural networks requires shifting from a dictionary of concepts to a geometry of representations; where meaning is encoded not in single atoms, but in the structure they collectively induce" 100% agreed.

@22loops @Strategy someone is going to liquidate them and this is going to be real

"MicroStrategy(@Strategy) sold 6,556 $BTC($555.85M) at an average price of $14,785 last week."

@teodorio we stayed in the main building recently 39th floor, the view was mental. you can actually take an elevator to the top floor, there is a restaurant, the view at night is crazy

Okura Prestige is the best hotel I've stayed in Japan hands down

@emm4x3 @NeelNanda5 can you share the link to his tutorial, please

Lately I've been craving a better mental model of transformer architectures, so I reimplemented GPT2 from scratch (credit to @NeelNanda5's incredible tutorial)! A lot of the magic comes from understanding the little details, and the more I learn, the more I realize that this is a field ripe for exploration. github.com/emma-x1/ml-fro…

incredible article and great insight there where you mentioned that it's harder to find and maintain signals that are less sensitive to latency - always had in my head that the best way to compete as an individual is to actually look for signals that are not so sensitive to latency listened once to a talk Alex Gerko gave, and he kept highlighting how he, in fact, focuses xtx on signals that are less latency sensitive - maybe it's an absolute ton of data that they have that helps there a lot

for example, if this statement is true: "when this bot puts capital on a side, that side is the eventual winner 95% of the time.", then it effectively gives us a formula for free money. we simply watch which side the bot takes, and if the price at which we can realize the trade ourselves is below 0.95, then we are expected to make profit. so something must be off. from the above, it seems like you can short sell sides, I thought that was not possible - maybe the above model does not account for the fact that you can mint both sides of the trade, i.e. it's possible to mint both YES and NO for $1.00. So if he is taker selling, then he must be selling the minted tokens. There are also neg risk tokens, which have their own mechanisms - iirc they actually might be enabled in sports markets.

@AlgorithmicBot let me know if you agree with this analysis

hmm. 95% accuracy across a bunch of different sports markets is very unlikely. was thinking this could be a latency play, where he observes the game feed faster than everyone else, but then he would be closing his positions as soon as the market trends to his fair price, unlike what supposedly this bot does, which holds to resolution?

@jonjonclark thanks! damn, I was once looking at this bot and thinking they must be doing something special, since they load up in size sometimes when the price for yes/no is super low (one of the strategies; not the only one) - but your write up here makes much more sense now

@AlgorithmicBot @rn1?tab=activity" target="_blank" rel="nofollow noopener">polymarket.com/@rn1?tab=activ… here they are smashing the orderbook every minute