lemursage

This is so weird, seeing this. Two years ago, I got a customs notice from FedEx asking to fill in my details. That was just a plain email from __some guy__ at FedEx with a PDF file attached. I wasn't expecting any package.

I wrote to their chatbot (of course, no human assistance) and after some time of "prompt engineering," or what one might call coercing, it finally directed me to a human consultant who confirmed it was indeed not a scam, and that it was indeed their messaging.

I opened the PDF, and it was pre-filled with someone else's data, with blank rectangles placed over fields in a bad attempt at redacting them (you could just move those rectangles around to reveal the underlying data).

The package later turned out to be a surprise from collaborators abroad. Years later, I still feel that scam aftertaste whenever I see the FedEx logo.

Reference on LLM architectures

It seems that the leaderboard doesn't contain the results for all of the supported DBs (I was looking for the pgvector myself).

The README.md contains a screenshot from local testing that's got more results included: https://github.com/zilliztech/VectorDBBench?tab=readme-ov-fi...

I'm "almost" a lifelong user of a cochlear implant. I got my first one when I was 9. Before I got it, I was communicating through lip reading and speaking, I never knew sign language. Lip reading I still use relatively often -- when I'm at a crowded restaurant, or at an unbearably noisy party, and there's many interlocutors at the table, I persistently stare at their lips. They take me for a great listener, when in fact, I can't hear shit, and I'm desperately switching back and forth between people's mouths to catch what they're saying. I'm out of shape and this takes so much of my brain power to understand people that I often cannot contribute my thoughts.

Though my cochlear isn't perfect, I would never think of not getting it. In fact, I'd probably be rather angry at my parents for not helping me get one as soon as it was possible. During my childhood and up until late college, I've only ever met one person who was so severely hard of hearing and was about my age, and that was where I have been getting my speech lessons before I got my first cochlear implant.

I really like the book's subchapter on colours, wish it was even more elaborated on. Colours are one of the subtle things I so often find difficult to get right.

As to using matplotlib in published research: when I started out as an undergrad, everybody in the research team used OriginLab for plotting -- my impression of it then was pretty good. At some point, I started using matplotlib + Latex + science plots and it caught on, mostly because there's no need to shift all the data around to a separate programme. Scienceplots package does heavy lifting with fonts and styling for specific journals, so it's just a matter of designing the right plot geometry and information density [1].

[1] an obligatory Tufte citation.

A timeless classic from Buena Vista Social Club comes to mind https://www.youtube.com/watch?v=o5cELP06Mik

In larger companies, and, specifically, bigger projects, systems tend to have multiple ML components, and those are usually a mix of large NN models and more classical (ML) algorithms, so you end up tweaking multiple parts at once. In my case optimising for such systems is ~90% of the work. For instance, can I make the model lighter or go faster and keep the performance? Or, can I make it go faster? Loss change, pruning, quantisation, dataset optimisation etc. -- most of the time I'm testing out those options & tweaking parameters. There is of course the deployment part, but this one is usually a quickie if your team has specific processes/pipelines for this. There's a checklist of what you must do while deploying, along with cost targets.

In my case, there are established processes and designated teams for cleaning & collecting data, but you still do a part of it yourself to provide guidelines. So, even though data is always a perpetual problem, I can shed off most of that boring stuff.

Ah, and of course you're not a real engineer if you don't spend at least 1-2% of your time explaining to other people (surprisingly often to a technical staff, but not ML-oriented) why doing X is a really bad idea. Or, just explaining how ML systems work with ill-fitted metaphors.

For a numerical "physicist" (yes, the quotation marks are indispensable), Sympy was somewhat of a godsend to me. Great for prototyping even more advanced models before optimising them later on in C++.

I haven't used Mathematica much, but I have a feeling that it's still more symbolically powerful (or requires less wrangling) than SymPy? I'd appreciate if somebody with more experience in Mathematica could lay it out flat for me if that's the case.

I wonder if this is in response to Sci-Hub/Arxiv proliferation or what else. Anyhow, I find it beyond rich -- the authors usually do all the typesetting, spell/grammar checks, formatting etc. themselves and now they get to pay for that (or their respective funding institutions). I presume that is what would land under a "processing fee". Beyond atrocious -- some journals have their custom LaTex templates only to discard formatting they put in them at the pre-publish stage. Also, reviewers are probably still not getting paid.

I'm not sure if that's an accurate sentiment, but I feel like there's an academic research winter with grants and funding are being slowly squeezed out of academia and moved to external consortia with loose academic affiliations (at least this is what I _think_ I see happening in Europe). If that's true, this trend of moving the paywall from readership to authorship is just sad.

As an engineer my concern about "google killers" is that I can't see an easy way to scale and control/optimize them in business settings. Apart from factual misstatements happening in the ChatGPT, what about source attribution? How is the relevance of a source determined? How is the flow of information through the network preserved (sourceA => sourceB => sourceC)? With Google we also don't know exactly but I can image some version of PageRank as tuneable. Finally, how to add new pages to index and measure potential "forgetting" that could happen?

Unless somebody could clarify those for me, this is what currently petrifies me -- some uncontrolled black box presenting its clandestine view of the web with no way to follow the breadcrumbs.

I'm a certified member of the Peripatetic school [1]. The ancients knew it and I know it -- I walk therefore I am. It's by the walking alone I set my mind to motion.

Needless to say that my downstair neighbours hate me for it. I guess philosophers were always ostracised ;)

[1] https://en.wikipedia.org/wiki/Peripatetic_school

Here's the catch -- I don't have a dog! Utter ramblings, I say!

I actually tested a bunch of those on GH Copilot -- sure enough, it does autocomplete certain prompts. No wonder, some are probably known quotes. Anyway, when I'm feeling down in my VSCode environment, I enter some random prompts in search of a pun. Most of the time it's just random rambling but hey:

# (prompt: my dog ...) seems to be interested in the plot.

# I wonder what the dog is thinking about.

# Probably about the food.

# I should feed him.

under some code where I was plotting a forecast.

Maybe many PhDs attend conferences just because they have to have some paper/poster presentation rather than they are genuinely interested in making new connections? That's what I'm guilty of at least.

I see a lot of folks touch on the UX of the Medium website. For me it's the content -- the paywall and overall layout would be digestible if not for the awful lot of generic Medium posts that regurgitate for the nth time the same basic tutorial of X, or repeat ad nauseam the particular set of ideas found on some independent blog sites.

I don't deny there's some quality content out there, but I feel it's not well curated and every other search phrase, I'm getting bombarded with tons of bland tutorial-blogposts (in ML domain especially) which really makes me disdain the website itself.

My grandma and her friends were projectionists in a small town cinema. It really did have a climate of the early scenes from Cinema Paradiso (although it wasn't that long time ago). I snuck into the balcony gallery for every movie I wanted to see. Though, I distinctly remember I came more for the atmosphere rather than the movies themselves. The cinema building was in a church-adjacent old theatre, over 250 years old. Before or after the film, I'd often hang out in the ticket booth, which was the cosiest place on Earth I know till today. Tiny, wooden enclosure lit with, I believe, sodium lamps. And of course, I had access to all the merchandise, particularly the posters. There were couple of old ones, and I feel like they don't really make them so artsy nowadays, so a bit of that poster magic eloped for me over time.

I think the problem here is that articles and publications in general do not provide enough detail to reproduce the results. Even though, that is the exact reason you write an article -- to make research reproducible. Peer-reviewed journals should, and they progressively try to be more strict about it, enforce the necessary level of detail and open-sourcing of methods.

Reproduction in science is a though problem -- not only it requires specialist lab equipment, methods or procedures, but sometimes "a trick" is needed. Inconveniently, it may not be mentioned in the publication for whatever reason. Therefore, giving an unexperienced PhD a task of reproducing somebody's work may end up to be a futile effort. But the idea is neat in general, I think.

Yes, « Nonlinear dynamics and chaos » is my favorite, I think it’s his best book. It has tons of nice examples and gets you a crash course over multiple concepts. « Sync » is nice too, but not a textbook definitely ( I don’t recall if it even had a single equation in it).

As an undergrad in EE and CS this was a bit of a mixed experience to me. Laplace transform was presented as kind of a nice hack, but I think the actual development of intuition is delegated to non math specific (engineering) subjects like for instance, Field Theory, Circuit Theory, Signal theory or Numerical methods. It seems that it’s expected that the students just blindly learn the basics of DEs which serve as one of sieves for non-math inclined students (here I may controversially say, rightly so!). You’re simply to swallow bunch of dry scones aka theorems and only then learn how to sue them.

I enjoy the overall fashion in which he presents the topics — clear, motivating example to start with, building on relevant details etc. His textbooks read like a good popsci book, yet they are of course much more detailed. I’m not sure how suitable his e.g. « Nonlinear dynamics and chaos » are for math students, since it feels a bit informal, but for developing intuition, interest and the scope of the subject, I haven’t seen anything better.

After reading Strogatz, I view past courses I have taken in engineering DEs as a crime on students and lecturers alike. I know there's a substantial difference between pure/applied maths courses and engineering, but a massive failure to make such a broad subject boring with dry, abstract examples is still beyond what I can forgive. That said, the courses did not traumatise me nearly enough to stay away from DEs which I guess is the only positive take I have at the moment.

It's not a statement about your general motivation for a PhD -- I don't think you can easily find a subject for which you will find a non-stop motivation 24h/7days for 3-4 years (or so, eqv. period required to get a PhD). If you encounter a dip in interest, which most likely will happen, you need to understand that changing a PhD is not always a best option -- just because it sounds good now, it doesn't mean it will in 2 years time. You supervisor should be there to guide you into interesting areas of research and may play a substantial role in rekindling your motivation. Naturally, mental health issues are another problem -- but I think they have often different origins than "just boredom" with subject. I may be completely wrong though.

« You will like any subject once you get to know enough about it. »

this is a really important observation — good attitude will do wonders to your motivation for research.

That was a really great listen, thanks! It actually mentions some of the concepts discussed around here (up or down, may be hard to tell in a while) about how people perceived the Commonwealth and the Republic.

I was also not quite aware of such feud! I generally assumed that Poles and Lithuanians got along pretty well, or at least that was my impression from talking to Poles. Some time later I asked two Lithuanian guys what's up with this, what I saw as, asymmetric feud, and they explained that it may be in part due to some communities in the Polish minority in Lithuania, which were rather pro-Russian. Needless to say I was told that many years before the war in Ukraine, way before 2014 (no idea how the perception is now). This really perplexed me to the point where I permanently gave up understanding this kind of feud/friendship generalisation -- not only this stuff is extremely local, region-wise, and in the long run, it just does not hold up.

One interesting thing is the perception of Jogalia aka Władysław II Jagiełło (King of Poland and Grand Duke of Lithuania) by Polish and Lithuanians nowadays. I'm pretty sure he's quite cherished historically in Poland due to his relatively stable rule and rather famous wars waged against Teutonic order, incl. battle of Tannenberg (the OG battle). On the other hand, it seems that there was a period when Jogalia was perceived in a rather negative light by Lithuanians. Not sure if 'traitor' is the right description of that sentiment, but definitely getting into a personal union, that ultimately led to the formation of the Polish-Lithuanian Commonwealth, seems to be judged rather unfavourably. From what I could understand, the dominance of Polish in that union was detrimental to the Lithuanian identity over time?

If anybody knows to what extent and for what exact reasons we see those differing pictures of the same person by peoples of two countries that constituted the same political body for quite some time, I'd be interested in hearing some details.

I found Katherine’s Ashenburg „Dirt on clean” book a nice read on how perception hygiene evolved over ages. It definitely reads as popscience, but for uninitiated ignorants like myself it was pretty sufficient. I see the author cites one of her other books in [6] as well.

The scope seems pretty standard but I really like strongly structured form and exercises. Is there a pdf version available perchance?

`Life is not short if you're lucky?*` I think that's what Seneca would have said back in the day. Nowadays, at least for the developed countries we could state `Life is not short if you're not unlucky`.

That said, many good things in life require time, work and careful planning (for somebody this could be building a house, finishing uni, perhaps raising a kid, tons of examples apply). Those that really bring fulfilment and satisfaction, where you can look back at this and say "I did that myself". Moment and day-like successes feel often like a serotonin hit, but do not persist long. So, I may say that I disagree with the premise of the article. What is often perceived as wasting time aka "watching Netflix", "playing games", "drinking with friends", turns out to be indispensable for us when we need to chew through mundane chores to reach the long-term goal we really strive for. But, if we're unlucky, we don't get to see the final result. A middle ground would be to start enjoying the small milestones?

This feels like that Futurama episode where Farnsworth discovers the last particle and descends into panic, realising he has nothing else to do. However, in my opinion, asking why there's nothing else is also a valid question to consider, even if we suspect there's something more to what we've seen so far.

Also, perhaps it'd be a tad bit more accurate to rephrase it `particle physicists`, not `physicists` -- even though it's totally exaggerated anyways.