nyrikki

71 ms·

421.
by nyrikki 2y ago
> A complex initiative requires picking the right tech that gives you a good benefit/cost ratio. That choice should never be a one way door, you don't control your customers or the future, it is the frame problem. The idea is t
422.
by nyrikki 2y ago
I disagree, those are all technical issues that are typically blocked by socio-technical and organizational problems. Retrofitting an API is a context specific problem, with few things that generalize but the above aren't the hard part
423.
by nyrikki 2y ago
> so there was basically no chance that models had been trained on submission code yet. I do suspect performance on AoC 24 will increase over time… No, but several of the problems were similar to problems in the training set from other c
424.
by nyrikki 2y ago
Looks like the site isn't responding, as it is from 2017 here is the archive.org link. https://web.archive.org/web/20240403055813/https://idea.popc... Previous discussion: https://news.yc
425.
by nyrikki 2y ago
> I would like to understand why numbers looks like leaves and not some other fractal like snowflakes for example (the inverse question, why biological leaves have this specific self-similar structure may yield more fruit). Snowflakes ar
426.
by nyrikki 2y ago
Protecting yourself from technologies, vendors, etc... are all very very worth making abstractions for. The silly "only after # Times" type rules stolen from the DRY for DRY reasons is part of the problem and far from the solution
427.
by nyrikki 2y ago
This study shows that controlling 0.1% may be enough. https://arxiv.org/abs/2410.13722v1 I have noticed some popular copied but incorrect leetcode examples leaking into the dataset. I suspect it depends on domain spe
428.
by nyrikki 2y ago
IMHO, the mamul form is even better for golf, F(n) in O(n log n)
429.
by nyrikki 2y ago
Yes, in this case PARITY is determining if the number of 1s in a binary input is odd or even It is an effect of the complex to unpack descriptive complexity class DLOGTIME-uniform TC0, which has AND, OR and MAJORITY gates. http:/&#x2F
430.
by nyrikki 2y ago
What LLMs can do is limited, they are superior to wet-wear in some tasks like finding and matching patterns in higher dimensional space, they are still fundamentally limited to a tiny class of problems outside of that pattern finding and ma
431.
by nyrikki 2y ago
The parameters have moved to a class method, still within the class, which is now public, but still taking a string and an intrinsic int, the string, as being passed as a parameter is already in the string pool or perhaps on the heap outsid
432.
by nyrikki 2y ago
It all doesn't matter, these are code fragments that are not in context of a full program. Lets look at the labels on these code blocks (in a chapter called "Chapter 2: Meaningful Names") > Listing 2-1 > Variables with
433.
by nyrikki 2y ago
In that example the constructor parameters were just moved, is there some cost in overloading the empty params? Original (labeled bad by the book): private void printGuessStatistics(char candidate, int count) { String number;
434.
by nyrikki 2y ago
Thank you for doing the work to find that for me. I still don't see that as: > "Lifting variables into “global” (object) state" It is simply the inversion and extraction method that is commonly used. The value of it is lo
435.
by nyrikki 2y ago
Ya looking into this more this is a react hyper correction to the Law of Demeter, related to language limits vs broad concepts. Not being a UI guy, I am not the one to say how universal this is, but the react context API seems to use stamp
436.
by nyrikki 2y ago
I don't remember anything about lifting state up in the clean series, was it perhaps a react specific book? To me that would violate the dependency inversion principle that most of the books leverage heavily. I know that some languages
437.
by nyrikki 2y ago
There tend to be two camps with the Uncle Bob franchise as I see it: Those that fall for the way he sells it, as the 'one true path', or are told to accept it as being so. Those who view it as an opinionated lens, with some sensib
438.
by nyrikki 2y ago
That is the dream of the Open–closed principle in SOLID and Postal's law like in RFC 1122. While these ideas are good defaults, unfortunately they lead to unmaintainable, monolithic, insecure code. One needs to make decisions on trade
439.
by nyrikki 2y ago
Neither TC0 nor uniform-TC0 are physically realizable, they are tools not physical devices. The default nonuniform circuits classes are allowed to have a different circuit per input size, the uniform types have unbounded fan-in Similar to h
440.
by nyrikki 2y ago
GPQA scores are mostly from pre-training, against content in the corpus. They have gone silent but look at the GPT4 technical report which calls this out. We are nowhere close to what Sam Altman calls AGI and transformers are still limited
441.
by nyrikki 2y ago
That is part of the problem, LLM coding assistants often code out what a human would often use a core library for, only including when required, like importing deque in Python to do a BFS. An example I helped someone with on AoC last week.
442.
by nyrikki 2y ago
Then you add in the Brouwer Hilbert mess that still causes challenges with talking about where the a priori assumption of the law of the excluded middle causes real problems with modern computational challenge. https://en.m.wikip
443.
by nyrikki 2y ago
1) have books like 'The Art of Programming' on my shelf, as AI seems to propagate solutions that are related to code golf more than robustness due to coverage in the corpus. 2) Force my self to look at existing code as abstract da
444.
Equivalence of the Frame and Halting Problems (2020) [pdf](philpapers.org)
1 points by nyrikki 2y ago|0 comments
445.
by nyrikki 2y ago
Never touched a CMX, with the EditDroid being the oldest I ever saw and it had a trackball/scrub knob. But the vi keybindings are directly from the LSI ADM 3A terminal, which is what Bill Joy wrote vi on. It had overloaded hjkl for the
446.
by nyrikki 2y ago
Most OS's use package managers these days, expecting ANY language to not run into major problems with that is just asking for trouble. Remember that Version SAT is NP-complete: https://research.swtch.com/version-sat Ho
447.
by nyrikki 2y ago
Venvs solve that problem, they don't cause it. You activate the venv per project to get out of the global namespace problem. As python x.nn versions are major with the nn part, it is like trying to mix c11 with c19.... things will brea
448.
by nyrikki 2y ago
venvs are namespace isolation, they are like containers. Even in huge monorepos you can just use something like a Makefile to produce a local venv using PHONY and add it to clean too This is how I actually test old versions of python, with
449.
by nyrikki 2y ago
For LLM's no, but the explanation is wrong also, it has nothing to do with 'false memories' and has everything to do with how LLMs work. Here is the paper. https://arxiv.org/abs/2401.11817 RAG and fine t
450.
by nyrikki 2y ago
That will be very common in the corpus, thus pattern matching and retrieval more than 'reasoning'. Not that isn't useful, just expected behavior for an LLM these days. I would be interested in the output, because it is almost

More ›