LolWolf

43 ms·

391.
by LolWolf 9y ago
Just change the root password once the account is enabled; this fixes the hole. sudo passwd -u root It's sad we have to do this, though.
392.
by LolWolf 9y ago
You're right that part of the problem is, indeed, theoretical (it remains to be shown that BPP ≠ BQP, even if it is widely believed to be true), but part of the problem is also physical (we need to build a model of BQP---i.e. a quant
393.
by LolWolf 9y ago
See the above thread given in [0]. It sounds weird, but they are talking about the strong CT thesis (e.g. that all realizable models are polynomial-time-reducible to BPP), not just the usual one (that no realizable model can solve haltin
394.
by LolWolf 9y ago
In this case the "strong Church-Turing thesis" (usually also known as the "extended Church-Turing thesis"[0]) is usually written as the following: > A probabilistic TM can efficiently[1] simulate any realistic model o
395.
by LolWolf 9y ago
Yay! That's awesome, I'm glad (this was a bit of a problem with last year's EE364A exam, actually, since the solver for a problem was too slow, so we decided to not include it). The benchmarks are quite impressive as well! An
396.
by LolWolf 9y ago
If the variable is constrained to be an integer, the problem becomes NP-hard-as-hell (in both a complexity sense and an approximation sense), sadly. It also stops being convex, so splitting-cone methods can't be directly applied and th
397.
by LolWolf 9y ago
Yeah! Another way of seeing it is that the derivative is a small (infinitesimal) perturbation around a region of interest: Any input that isn't maximal will be some finite distance away from the maximum, so any small enough perturbatio
398.
by LolWolf 9y ago
This is pretty cool; Boyd had mentioned to me something about there being a good method for solving QPs that was nicer than the usual SOCP reduction (that's horrible and pretty slow in the general case). I think this was the case he wa
399.
by LolWolf 9y ago
I think it's not weight-by-weight equivalent. I believe that a tiny amount of Erythritol is enough to be equivalently sweet relative to a much larger amount of sugar. That being said: I'm not sure if it's enough to make it ch
400.
by LolWolf 9y ago
This would’ve literally been my answer. Brilliant. I can’t recommend this answer enough.
401.
Network connectivity correlates of variability in fluid intelligence performance(sciencedirect.com)
2 points by LolWolf 9y ago|0 comments
402.
by LolWolf 9y ago
But the claim the paper is making fits perfectly with the above: > On the negative side, we prove that no algorithm can achieve accuracy of ε < 2η in learning any non-trivial class of functions. Though this is for a very particular ty
403.
by LolWolf 9y ago
Not only do I think (b) is more likely than (a), I think (b) is strictly true. I mean, our concentration is quite limited and there are optical illusions which are, perhaps even more strongly than eps-adversarial examples for a given archit
404.
by LolWolf 9y ago
I'm not sure it's about how shallow the network is; even logistic regression can be fooled by the same techniques (e.g. 1-layer NN). That being said, maybe it does have something to do with linearity (I suspect not) or maybe it&#x
405.
by LolWolf 9y ago
> Although our models are more vulnerable to white-box FGSM samples compared to the v3adv model, ensemble adversarial training significantly increases robustness to black-box attacks that transfer FGSM samples crafted on the holdout Ince
406.
by LolWolf 9y ago
Sure, you can't claim that every machine learning system is susceptible to this attack (here's one that isn't, a model that always returns the same class on all inputs). But (apart from obvious trivialities) I suspect this
407.
by LolWolf 9y ago
It’s a final paper, not a scientific publication nor necessarily original research.
408.
Optimizers, momentum, and cooling schedules(guille.site)
3 points by LolWolf 9y ago|0 comments
409.
Some thoughts on global path optimization for drones(guille.site)
1 points by LolWolf 9y ago|0 comments
410.
Communication complexity of approximate Nash equilibria(arxiv.org)
1 points by LolWolf 9y ago|0 comments
411.
by LolWolf 9y ago
I've seen this paper enough to warrant a comment. The paper makes no reduction that the market is computing in P, so it's unclear to me how this conclusion of P=NP iff EMH is warranted by the paper. Perhaps more clearly showing th
412.
Game semantics(en.wikipedia.org)
1 points by LolWolf 9y ago|0 comments
413.
by LolWolf 9y ago
> But there is no conspiracy; that would presume far more competency than they have hired for. This seems to be the case almost universally, but I guess it's easier to jump to the 'malice' conclusion than to attribute it t
414.
Least-squares and image processing(guille.site)
2 points by LolWolf 9y ago|0 comments
415.
by LolWolf 9y ago
Oh, interesting! I'm so used to reading it in papers that they're first anesthetized before being put in CO2 (along with many lab experiment that do it) that I thought that was SOP. Thanks!
416.
by LolWolf 9y ago
You're right, but SOP for euthanizing rodents is CO2 after anesthetizing them.
417.
by LolWolf 9y ago
Sure, two partial functions are equivalent over a set d iff they halt on the same subset of inputs and have equal outputs whenever each terminates. Then, if you could decide equality of two functions, f, g, over a finite domain d, then let
418.
by LolWolf 9y ago
It works in the sense that there are finite number of them on a finite-size machine, but exhaustion would be infeasible in almost all possible cases.
419.
by LolWolf 9y ago
Category theory is fine, I guess, but there's no need to invoke a sledge-hammer when talking about 99% of most work here. It's a little like saying, yes, fine we can reference the hyperreals to derive infinitesimal calculus in a b
420.
by LolWolf 9y ago
Awesome, thank you!

More ›