kqr

21 ms·

481.
by kqr 11mo ago
For beginners to SPC, I wrote a practitioner's guide a couple of years ago. Might provide meaningful context: https://entropicthoughts.com/statistical-process-control-a-p...
482.
by kqr 11mo ago
What confuses me about deep nets is that there's rarely enough signal to be able to meaningfully train a large number of parameters. Surely 99 % of those parameters are either (a) incredibly unstable, or (b) correlate perfectly with ot
483.
Statistical Process Control: A Practitioner's Guide (2022)(entropicthoughts.com)
2 points by kqr 11mo ago|0 comments
484.
by kqr 11mo ago
How uh, does one find out about battery problems? I almost exclusively use laptops, and I tend to leave them plugged in most of the time. I don't want a sudden lithium-ion battery fire. Can I detect ahead of time that things are going
485.
by kqr 11mo ago
> The other extreme, where you can go inside everywhere, turns out to be boring But that's the point! Daggerfall is like this too: huge areas (both cities and landscapes) with nothing interesting in them. That's what makes them
486.
Denis Shiryaev's Quick Math Challenge(shir-man.com)
1 points by kqr 11mo ago|0 comments
487.
by kqr 11mo ago
> The deep feedback-loop analysis Thinking the consideration of feedback loops requires "deep" analysis is, I suspect, part of the problem! The insufficient feedback shows up at a very shallow level.
488.
by kqr 11mo ago
Thanks to the magic of compounding, inverting overbetting also leads to overbetting. Especially once costs are accounted for.
489.
by kqr 11mo ago
In defense of your taste, it was updated based on the loud feedback here, so you probably read a slightly better version than that which was flagged.
490.
by kqr 11mo ago
To be clear, I'm not expecting a full analysis within hours. I'm hoping for a method of analysis by which the major deficiencies come up at a high level, and then as more effort is spent on it, more details around those deficienci
491.
by kqr 11mo ago
I would be worried if they had all the answers within a few hours! I was just caught off guard by the focus on technical control measures when there seems to have been fairly obvious problems with information channels. For example, "mo
492.
by kqr 11mo ago
Maybe I'm misunderstanding something but it being a blocking call does not make timeouts less important -- if anything they become more important!
493.
by kqr 11mo ago
Part of my argument in the article is that it does't take long to come to that realisation if using the right methods. It would absolutely have been possible to identify the problem of missing feedback by that time.
494.
by kqr 11mo ago
> how these ideas were actively used by the author at e.g. Tradera or Loop54. This would be preferable, of course. Unfortunately both organisations were rather secretive about their technical and social deficiencies and I don't want
495.
by kqr 11mo ago
The article makes no claim about the effort that has gone into the analysis. You can apply a lot of effort and still only produce a shallow analysis. If the analysis has not uncovered the feedback problems (even with large effort, or withou
496.
by kqr 11mo ago
Yes, but they were either source package managers or they didn't resolve dependencies transitively, only complained about missing direct dependencies.
497.
by kqr 11mo ago
Note that 10th of October is before the trading period in this experiment. If anything, autoregression over shorter timescales would suggest entering after 10th of October being a good idea!
498.
by kqr 11mo ago
Ah, yes, of course. I must have missed that. Thanks!
499.
by kqr 11mo ago
One of the remediations listed is > Eliminating the ability for core dumps or other error reports to overwhelm system resources but this is not mentioned at all in the timeline above. My best guess would be that the process got stuck in
500.
by kqr 11mo ago
No, LLMs are not a good choice for this – as the results show! If I had to guess, they're experimenting with LLMs for publicity.
501.
by kqr 11mo ago
I agree. Plus it's way too short a timeframe to evaluate any trading activity seriously. But I still think the experiment is interesting because it gives us insight into how LLMs approach risk management, and what effects on that we ca
502.
by kqr 11mo ago
Super interesting! You can click the "live" link in the header to see how they performed over time. The (geometric) average result at the end seems to be that the LLMs are down 35 % from their initial capital – and they got there
503.
by kqr 1y ago
Something like this: https://entropicthoughts.com/fizzbuzz-through-monoids
504.
by kqr 1y ago
How is this without conditionals?
505.
by kqr 1y ago
If it's any guidance, US cyber risk insurance (which covers among other things disruptions due to supplier outages) has continuously dropped in price since Q1 2023, with a handful of percent per year. If you excuse the sloppy plot manu
506.
by kqr 1y ago
Well, if you do get around to it, I might just be your first user!
507.
by kqr 1y ago
> xmonad is really really good. Especially how it handles workspaces on multiple monitors This is the reaon I use XMonad, and I hate how this one design choice that only one window manager has gotten right is what locks me into X11. If
508.
by kqr 1y ago
> it's definitely the case for Niri No, it doesn't seem that way. The Niri manual says each monitor has an independent set of workspaces. In XMonad there is one shared set for all monitors, but each monitor independently shows
509.
by kqr 1y ago
> Guess you hacked the Plot.ly widget? Paintbrush in Gimp on top of screenshot.
510.
by kqr 1y ago
Good catch. If there are n samples, and the lower half of them are equal to or less than 0.29, then a total mean of 0.82 would require that sum of the the upper half must be greater than 0.82n - 0.29n/2 = 0.675n. For n/2 numbers t

More ›