dualmindblade [he/him]

  • 7 Posts
  • 114 Comments
Joined 4 years ago
cake
Cake day: September 21st, 2020

help-circle











  • So you have some kind of spectrum modem with a built in router and the wired connection is solid? Assuming you’ve turned it off and back on again…

    Last time I had spectrum you could still configure quite a bit from the modem admin UI, usually you can access it at 192.167.1.1, though this can vary. Get in there, if you changed the password (you should have lol) and don’t remember you’ll need to do a hard reset on the modem which you can look up by model number. Update the firmware and/or mess with random wifi settings till it works. All of this is reversible with a hard reset, unless the firmware update fails so maybe do that last. If this fails either replace the hardware or get a separate wifi router, don’t need to worry about compatibility really just get one with the same or greater Ethernet bandwidth as the model. Or you can go to the spectrum brick/mortar and wait in line and they’ll replace it.

    VERY IMORTANT: when you do end up getting off spectrum and have to return the hardware, take a picture of the receipt at UPS or wherever you happen to return it. They will very likely try to fuck you and say it wasn’t returned, I believe this is an unofficial internal policy. Even if you do this they may send you to collection even thought you proved to them more than once you returned it. Take screenshots and/or recordings of all interaction with spectrum agents. Then finally a strongly worded letter (registered mail) to the collection agency with all your evidence will end the nightmare. Is it worth it to spent 12 hours of work and $10 to send a registered letter so that spectrum only extracts $2 extra dollars by selling your “debt” rather than 50? I would argue that yes, it is











  • It really is, another thing I find remarkable is that all the magic vectors (features) were produced automatically without looking at the actual output of the model, only activations in a middle layer of the network, and using a loss function that is purely geometric in nature, it has no idea the meaning of the various features it is discovering.

    And the fact that this works seems to confirm, or at least almost confirm, a non trivial fact about how transformers do what they do. I always like to point out that we know more about the workings of the human brain than we do about the neural networks we have ourselves created. Probably still true, but this makes me optimistic we’ll at least cross that very low bar in the near future.