stared

47 ms·

871.
by stared 2y ago
Once you go beyond Markov chains and into vector models (Transformers or even LSTMs), next —-token-- syllable prediction can capture grammar.
872.
by stared 2y ago
I am a big fan of Divvy ( https://mizage.com/divvy/)—while both dated and paid, it has a wonderful feature of a visual tile map you can use. In all other cases I know, there are a bunch of shortcuts to make a window upp
873.
I asked the Dalai Lama about self-hatred(wisdom2be.com)
2 points by stared 2y ago|0 comments
874.
by stared 2y ago
I never understood the randomness argument against art. Wasn't there a lot of randomness in Pollock's use of paint on his canvases? Also, one can extend this thinking to argue that gardens (or mazes) cannot be a piece of art since
875.
by stared 2y ago
Well, see my current work in progress: https://github.com/stared/balfolk-viz/blob/main/src/componen... and D3 part in https://github.com/stared/balfolk-viz/blob/main&#
876.
by stared 2y ago
From "Kolmogorov Complicity And The Parable Of Lightning" ( https://slatestarcodex.com/2017/10/23/kolmogorov-complicity-... ): "Did Giordano Bruno die for his astronomical discoveries or his athe
877.
by stared 2y ago
Well, D3 is not a set of tools for charts (alike matplotlib, ggplot or so), but a low-level library to create your data visualization from scratch. Depending on your goal, it may or may not be what you need.
878.
by stared 2y ago
I've been in love with D3.js and have used it as the key technology for dozens of data visualization projects, both commercial and personal. However, its rendering syntax now feels dated, reminiscent of the jQuery era. While I still us
879.
by stared 2y ago
Except when something takes a different time than needed, it makes a domino effect on the calendar. And even if something takes less time, it has a disruptive effect on the calendar. Writing a piece of documentation (no code, nothing can co
880.
by stared 2y ago
See Seskir et al., "Quantum games and interactive tools for quantum technologies outreach and education", Optical Engineering 61(8), 081809(2022), https://doi.org/10.1117/1.OE.61.8.081809 (open access, but if
881.
by stared 2y ago
I created a list of Interactive Machine Learning, Deep Learning, and Statistics websites ( https://p.migdal.pl/interactive-machine-learning-list/ ), which includes Immersive Linear Algebra, among other things. For a broa
882.
Effort – a possibly new algorithm for LLM Inference(kolinko.github.io)
130 points by stared 3y ago|4 comments
883.
by stared 3y ago
> I reckon even a Markov chain model of similar magnitude could wow us in various ways. I bet $100 it cannot.
884.
by stared 3y ago
Extraordinary claims require extraordinary evidence. Otherwise, it is a clickbait (if done intentionally) or delusion (if not). Well, I was shocked to see LLMs (rather than something intrinsically related to Reinforcement Learning) reach th
885.
by stared 3y ago
It is not even remotely up to date. > Cursor (editor, not yet released) Well, it is up and running, https://cursor.sh/ . I recommend it a lot - finally, not just a chat, but a tool in which a considerable fraction of prog
886.
by stared 3y ago
It is a corollary. Still, if the human remains are an issue, well - in numerous ritual burials of rulers, there are remains of servants or concubines. In most cases, we can assume it was not consensual. In this case - should we cremate (or
887.
by stared 3y ago
So, the solution is to remove it from public access so it cannot be used in disrespectful practices. Not to destroy the artifact. Especially not by a university or any institution supposedly preserving knowledge and culture. (And yes, most
888.
by stared 3y ago
By this reasoning, we should destroy all buildings that were constructed with slave labor; and all jewelry with gems acquired by similar means.
889.
by stared 3y ago
I like saying that it is an S4 - set, setting, substance, and soul. And the substance (as long as it is a tryptamine) is less important, save for personal preferences. (I have heard a lot about LSD for fun and shrooms for insights - for me,
890.
Every Tangled Cord in Your Junk Drawer Ranked(hard-drive.net)
2 points by stared 3y ago|0 comments
891.
by stared 3y ago
Actually, a handgun is unlikely to kill bacteria.
892.
by stared 3y ago
I am curious: What is Julia's current niche? I mean, I tried it a few times a decade ago, but the environment was not mature, and performance improvements were not in the places I needed. Though, I guess, things moved from there. Pytho
893.
by stared 3y ago
People in the industry think that doctoral advisors are something like bosses in companies. They aren't. A Ph.D. is an all-or-nothing experience. When leaving on bad terms, you not only lose a few years of your life, but it is likely t
894.
by stared 3y ago
In my case, relating is what makes me felt listened and understood. I feel that by "relating" many people mean "comparing" - and this one is annoying. At the same time, I cannot stand "it must be hard" or simil
895.
by stared 3y ago
There is already a body of research on the perception of time as a function of age, see https://en.wikipedia.org/wiki/Time_perception#Changes_with_a... . It stems from a few things, and novelty is only one factor. (Also
896.
by stared 3y ago
A fair point - that one thing is the absolute skill level, and the other is a comparison with prior expectations. The latter may give a positive surprise or deep disappointment.
897.
by stared 3y ago
Yes and no. I mean, a lot of moderates would like to avoid seeing any extreme content, regardless of whether it is too much left, right, or just in a non-political uncanny valley. While the Horseshoe Theory has some merits (e.g., both left
898.
by stared 3y ago
It is not only about morals but the incentives of parties. The need for sexual-explicit content is bigger than, say, for niche artistic experiments of geometrical living cupboards owned by a cybernetic dragon. Stability AI, very understanda
899.
by stared 3y ago
It is hard to define a fair comparison between humans and LLMs. Artificial "neurons" (i.e., vector entries + ReLU) have little in common with biological neurons (each one is a computing unit on its own). Data efficiency is crucial
900.
by stared 3y ago
I agree, the lack of consistency sets us (i.e. humans vs LLMs) apart. > LLMs are confidently and repeatedly wrong about very simple things that a human would never be. It is a strong assertion. A substantial fraction of the population (m

More ›