zburatorul

Boltzmann constant is not fundamental. There's no need for it at all if we switch from measuring temperature in Kelvin to Joules.

The Higgs.

I can see how scale free systems have their action stay invariant under more transformations. I'd like to better understand the connection with action stationarity/extreme. Can you say more?

Fidelity supports ofx downloads which makes the process easier than clicking buttons in a browser. OFX is very detailed, and ofx to cab conversion is solved I think.

I highly recommend this service as well.

This should be the top comment in this thread imo. This service solves a lot of the issues mentioned. No need to call anyone and beg for a cancellation, just close the card.

A caveat is that one should be aware of how binding one's agreement with a business is, and whether a simple card cancellation will make the problem go away or become worse (collections).

And I highly advise caution with TokenTax. I chose them because I had ~2k transactions on Deribit, and they promised they can digest them. They don't give out refunds specifically because their website claims they will work with you until you're satisfied.

Well, it was clear from the 5th minute they don't have "Deribit support", and I had many email exchanges with them spelling out in detail how their engine misunderstands my transactions (cost basis etc.). I gave them every opportunity to improve, and instead they eventually just ghosted me.

I understand it may be unrealistic to expect they will build this capability in three weeks. It's a shame they weren't able to level with me, and talk about the complexity of this. I ended up writing a taxable gains calculator for Deribit myself and would be happy to share it with them.

This sounds interesting. Can you share more please? It sounds like there is some multilingual input text on the basis of which you make recommendations, but I think you would have called that a search engine rather than recommender.

So I'm using the gradients "grad" of the second network (the DDSP) to make a loss function for the first network, such as |grad|^2?

Do you think DDSP could be used for feature engineering/discovery? My setting is a time series that I want to do regression on. But it's not clear a priory what features of the series have good predictive power. I imagine making the DDSP the first layer of a FNN and the gradients helping me identify the right filters to use to extract important features from my data.