mrfox321

10 ms·

31.
by mrfox321 3y ago
Yup I've had recursive dreams. You're likely internalizing something you're grokking
32.
by mrfox321 3y ago
I'm not saying it's 50/50. But future signees are influenced by previous signees. Acting in good faith is different from bias.
33.
by mrfox321 3y ago
The key is that the support is not anonymous.
34.
by mrfox321 3y ago
Doing the math, it is extremely unlikely for a lot of coin flips to skew from the weight of the coin. To that end, observing unanimous behavior may imply some bias. Here, it could be people fearing being a part of the minority. The minorit
35.
by mrfox321 3y ago
What would you do if 999 employees openly signed a letter and you are the remaining holdout.
36.
by mrfox321 3y ago
Lexus is Toyota. In Japan, they are branded as Toyota, hood symbol and all. They just market them differently in the West. The same thing goes for the Acura / Honda distinction.
37.
by mrfox321 3y ago
Your intuition on search being implemented is correct. It's still TBD on whether these new generations of language models will democratize search on bespoke corpuses. There's going to be a lot of arbitrary alchemy and tribal knowl
38.
by mrfox321 3y ago
It can be stated as follows: The number of documents (denoted as N) to search over consumes resources and increases the overall latency of the search infrastructure. However, the amount of traffic ebbs and flows. Under periods of lower tr
39.
by mrfox321 3y ago
You don't want faster solutions? Even when you really need to value your time?
40.
by mrfox321 3y ago
When you say hashed bags of sub graphs, are you generalizing bag of words, here?
41.
by mrfox321 3y ago
If your system cannot be decomposed away from shared mutable state, then you cannot avoid lifetime management and synchronization primitives. Ultimately, it depends on your data model.
42.
by mrfox321 3y ago
because they are assuming the reader has already built a query engine and is now re-writing it. first prototype: bottom-up design second pass: top-down design
43.
by mrfox321 3y ago
That sounds intractable. How would you differentiate mojo code from vanilla python without a ton of boilerplate at language boundaries.
44.
by mrfox321 3y ago
The same group think happened during COVID. Tech closed up shop on a Wednesday. Finance shut down the following Monday.
45.
by mrfox321 3y ago
I think you overestimate the likelihood of being judged on code in GitHub.
46.
by mrfox321 3y ago
Typo in cross entropy, should be p log q
47.
by mrfox321 3y ago
Yeah matrices are more abstract. I guess I am just pointing out that your concrete example of non-commutative rings (matrices of matrices) still needs a proof to demonstrate bijection between 4N x 4N (scalar) and 4 x 4 (N x N(scalar)). Blo
48.
by mrfox321 3y ago
Re [1]: it's fairly concrete to simply say that matrix multiplication can be performed block-wise.
49.
by mrfox321 3y ago
Are you aware that the gzip paper fudged their accuracy numbers by assuming an oracle could correctly pick from the nearest 2 neighbors with 100% accuracy? In other words, they published top-1 accuracy from top-2 accuracy calculations. I wo
50.
by mrfox321 3y ago
You are misreading things. Regardless of numerical stability tricks (e.g. exp(x_i-max(x))), you are still simply normalizing the logits such that the probabilities sum to 1. The blog adds an additional hidden logit (equal to 0) to allow for
51.
by mrfox321 3y ago
Einstein expanded several fields of physics. Special + general relativity Microscopic definition of Brownian motion Photoelectric effect Mass-energy equivalence
52.
by mrfox321 3y ago
Sure it does. Have you seen the type system used to generically dispatch matrices to GPUs or cpus. Or the dispatching to give you auto-differentiation? Maybe it's not your definition of "new ideas", but they are really useful
53.
by mrfox321 3y ago
Right, so it's less efficient than async. Async would let you yield at the gather.
54.
by mrfox321 3y ago
Because "Instagram is cool" Joking aside, this is a branding decision.
55.
by mrfox321 3y ago
Isn't that the opposite? That is, bypassing user space, not kernel space?
56.
by mrfox321 3y ago
Deflation can be self-reinforcing, too. If both directions can lead to runaway feedback, it likely needs to be controlled.
57.
by mrfox321 3y ago
I'm not in the field, but some of your questions are research problems regarding boundary conditions of the laplacian operator. There's also similar work framed in terms of billiards: https://en.m.wikipedia.org/wik
58.
by mrfox321 3y ago
same with python: @functools.cache def fn(x): https://docs.python.org/3/library/functools.html#functools.c...
59.
by mrfox321 3y ago
Yes, that's the definition of aperiodic tiling. There is no translational symmetry.
60.
by mrfox321 3y ago
It does not plateau at 100mcg...

More ›