@Stoiiic I think I probably misunderstood the initial post. I thought you were using the asset z score against the whole universe, not that each z score is asset specific
@banteg Claude finds issues with itself as well. Try it the other way around and claude will find issues with codex. The problem is that all models, in my experience, tend to even create issues and mess up your code, if you even slightly suggest that finding one is the "goal"
@zhusu Well, the most known fact in trading is that far more money has been lost buying highs than any other type of strategy, and it's not even close. You should know something about that
@pedma7 You need to do the "divide and conquer" part by yourself. All models I've tried perform significantly better with smaller parts of code and in small chat logs. Also, try Claude, it's currently much better and code is much cleaner imho
@tomwanhh@ether_fi Thank you for your work, I found your dashboard to be very useful. Any chance you can add a row with the summaries of the numerical data on the table? Also a graph of "Change" by Allocation size group would be very interesting to see
@InuraNetwork Launch the same token, on a different chain at 1/30th of the price at the moment and then proceed to market dump 10eth straight from deployer wallet for like 60% price impact. BASED.
I swear crypto people don’t even think through the first-order impacts of their ~ideas about the future~ nevermind downstream ramifications. Some of these ppl are ‘professional investors’ managing half a bil of other ppls money and their thesis is “bro… wouldnt it be cool if…”