thesz

28 ms·

481.
by thesz 2y ago
Parquet is underdesigned. Some parts of it do not scale well. I believe that Parquet files have rather monolithic metadata at the end and it has 4G max size limit. 600 columns (it is realistic, believe me), and we are at slightly less than
482.
by thesz 2y ago
The fear of the world around makes one less mobile. One is too afraid to go out, basically. The ability to be mobile is what makes depression to go out. [1] https://www.ncbi.nlm.nih.gov/books/NBK99429/ [2] htt
483.
by thesz 2y ago
HyperLogLog uses additions, it keeps sums. Thus, you can subtract one HLL sums from other. This is useful if stream supports deletion. Streams with deletions can be found in log-structured merge trees, for one example, so one can estimate c
484.
by thesz 2y ago
Page 15 of KAN paper deals with Feynman datasets.
485.
by thesz 2y ago
I once explored the Epigram dependently typed programming language. It used to preclude many types of free form pattern matches, due to heavy dependence on structured editor (you speciy a type, it generates pattern matching, you cannot chan
486.
by thesz 2y ago
"Haskell is the best imperative language," (C) various software engineers. Also, algebraic data types can be seen as hierarchy consisting of abstract base class and several final children classes. So it is an inheritance model, ju
487.
by thesz 2y ago
The paper does not cite [1], which describes prediction by partial match algorithm with unbounded context length. [1] https://www.researchgate.net/publication/2473004_Unbounded_L... That is from 2003 and actually quite
488.
by thesz 2y ago
https://gradientdescending.com/unsupervised-random-forest-ex... You can cluster data using unsupervised random forests and then use these cluster indices as features.
489.
by thesz 2y ago
Note that KANs use LBFGS, which is second-order optimization method. My experience with the use of second-order features suggests that simple gradient descent often leads to divergence.
490.
by thesz 2y ago
https://arxiv.org/abs/1210.7273 In the article "Distilling free-form natural laws from experimental data", Schmidt and Lipson introduced the idea that free-form natural laws can be learned from experiment
491.
by thesz 3y ago
How does it compare to fractal video compression [1]? [1] https://www.computer.org/csdl/proceedings-article/dcc/1992/0...
492.
by thesz 3y ago
You cannot find use of decision tree (aka "bunch of IFs") deduction and use in the current field of AI. You can find use of neural networks training and inference in the current field of machine learning.
493.
by thesz 3y ago
The most interesting part there is that Bayer is a pharmaceutical company. They should have chains of responsibility and self-organizing these chains can be... tricky if not impossible. Mostly due to probability to be a plaintiff in a lawsu
494.
by thesz 3y ago
Before LC there were combinators [1]. [1] https://en.wikipedia.org/wiki/Combinatory_logic These are even more primitive.
495.
by thesz 3y ago
Usually, LLM's output gets passed through beam search [1] which is as symbolic as one can get. [1] https://www.width.ai/post/what-is-beam-search It is possible to even have 3-gram model to output better text predi
496.
by thesz 3y ago
I guess it is funding. Compare funding of Google+Meta to what was/is available to Cyc. Cyc was able to produce an impact, I keep pointing to MathCraft [1] which, at 2017, did not have a rival in the neural AI. [1] https://en
497.
by thesz 3y ago
There are lattice-based RNNs applied as language models. In fact, if you have a graph and a path-weighting model (RNN, TDCNN or Transformer), you can use beam search to evaluate paths through graphs.
498.
by thesz 3y ago
Lenat was able to produce superhuman performing AI in the early 1980s [1]. [1] https://voidfarer.livejournal.com/623.html You can label it "bad idea" but you can't bring LLMs back in time.
499.
by thesz 3y ago
I do not write Rust. My day-to-day job is C++'s application support and, frankly, what Rust offers does not help there. When I do programming for fun, I prefer to use Haskell into which "Rust" can be embedded to, if I need th
500.
by thesz 3y ago
> Having a cheap laser able to cut metal at home / small shops would be so useful (And sneakily dangerous as lasers are). So, please do not come around these lasers with remaining eye unprotected. I think one should work with these
501.
by thesz 3y ago
It should be named "I accidentally a meme search engine" [1]. [1] https://www.reddit.com/r/AskReddit/comments/jooo5/reddit_ori...
502.
by thesz 3y ago
> Despite a remarkable amount of algorithmic effort on Boolean satisfiability (SAT) and related problems, to date no one has invented an algorithm with faster-than-exponential (O(2^n)) running time; indeed, there is no polynomial-time al
503.
by thesz 3y ago
The LLM under test was GPT 4 turbo.
504.
by thesz 3y ago
"Please implement blocked clause decomposition in Haskell" I, actually, asked my friend to ask whatever LLM he paid for to perform this experiment. I think he uses latest or previous GPT. At least he is paying for it which I canno
505.
by thesz 3y ago
Blocked clause decomposition is implemented in, I think, two SAT preprocessors in SAT solvers, both are C++. There are about three to five different articles on the topic. For LLM to implement BCD in Haskell it needs to be able to translate
506.
by thesz 3y ago
I will include such disclaimer after you prove me wrong as you have means to do that.
507.
by thesz 3y ago
London taxi drivers that need to remember approximately 35 thousands streets and their more or less recent state, have two changes in their brains: they have a dedicated brain area that helps with navigation and their hippocampus is bigger
508.
by thesz 3y ago
> a good model Strongly reminds me of "no true scotsman." If you have an access to a good model, try that simple problem of blocked clause decomposition and please get back to us.
509.
by thesz 3y ago
What you do with LLM will affect other persons, especially if the result is code. The cumulative time others spent reading your code will be much longer than you writing it. You go hang up with your five-ten friends and dozens of other men
510.
by thesz 3y ago
I think that LSP is a Language Server Protocol. I do not use that. I can understand its utility but I do not use it. "Compiler error messages as I type" distract me from, well, typing. I can pause to think and then I get an error

More ›