I don't think thats necessarily true, they aren't really capacity constrained in practice (they might be behind the scenes and adjust training on the fly, but thats speculation), so wasting tokens effectively helps utilize their (potentially idle) inference GPU's
Removing the "Market expert" which uses OHLCV (Open, High, Low, Close, Volume) also drops the sharpee from 5.01 to 1.88 while also increasing the max draw down to 13.29% (v.s. 9.70% for the index). I'd be very surprised if the pre training of the base model was the only source of leakage...
There is https://www.rwkv.com/ which is an LLM based on RNN's, thus having "infinite" context length, it comes with its own tradebacks though. (Notably that its impossible to actually store infinite information in the network, so it prunes based on which information it finds more important.)
https://x.com/tayvano_/status/1847877011462901915
This thread has some info about very similar past attacks, should give some insights into the level of sophistication that goes into something like that.
So now I probabilistically spam a ton of different orders to on average get my desired fill...
This just turns it into a "whoever is best at DoS'ing the exchange" game.
As the orderbook fills with competitor orders it makes sense for yourself to also spam orders so each of your orders maintains the same probability of being filled
I mean a generation of bees takes atleast a year since only the queen is really reproducing, the worker bees themselves aren't exposed to any evolutionary pressure since they can't reproduce. (I guess they are by proxy, once their queen dies so does their lineage)
> I wish there existed a modern cross-platform open source reactive GUI framework which was less of a resource hog than Electron
Flutter (Desktop) comes pretty close to that. Its really fast to write and your code works on IOS/Android/MacOS/Windows/Linux unless you need Platform specific API's, and even that is possible.