@BorisBarbour Interesting stuff! This highlights why it's so important to include the versions of all the software used for analysis. I find it so frustrating that this does not seem to be the standard across the board.
Even more egregious, some researchers don't even version their own code. I've spent quite a bit of time trying to convince them that this is critical. It's crazy to me!
@ChristosArgyrop I'm far from being an expert and very happy to be corrected on this... I think the main issue with this type of benchmarks is that whichever you look at will somehow be skewed.
Yes, <your favourite LLM here> can pass the bar exam, get an accreditation for becoming an airline pilot and create Michelin Star winning recipes... all while also being able to be convinced that 2 + 2 = 5.
Given that LLMs are tools, they also need a skilled person to use them. The AI industry loves telling us that you can ask anything to an LLM and you'll get perfect results, 99.9% of time. They forgot to mention you need to know how/what to ask, which means that with good domain knowledge you can create prompts where these tools fail miserably.
This is the same as using a search engine. If you don't know what to ask you will get irrelevant results. And if you don't have the critical skills to understand whether a reply is relevant or not... you will believe whatever you read.
Here is a quick #tutorial on how to build a #NeuralNetwork from scratch using #NumPy:
🌍 https://www.fabriziomusacchio.com/blog/2024-02-25-ann_from_scratch_using_numpy/
Politics
@GottaLaff @atrupar These are the same tactics that Berlusconi used in Italy during his time. Normalize what used to be shocking. Make people focus on some of his shocking acts (underage sex parties, for which I know some part of the population was actually sickly proud of him) to distract from the larger criminal scheme he was undertaking. All of this while paving the way for the fascist party that is in power at the moment, and leaving behind a trail of bigotry, misogyny, and corruption.
I never want to get to a point where I have to turn down random meetings with people asking for help with stuff. I take a meeting or two a week thats just like "how do I do this thing" or "how are you thinking about this" and they are often my favorite part of the (work) week. I have seen people on here host "open office hours" in a more structured way where they set aside an hour or two just for helping ppl with stuff, and I love that.
Thats one of those big hidden harms of overworked profs having to tend grants all the time, all that accumulated wisdom and no time to share it.
This is the single best article on color spaces I have seen around. Well written, well informed and illustrated with interactive visualizations.
If you only read one, this should be it. Truly brilliant work @eeeps!
There's a lot of speculation about whether OpenAI's video generation model #Sora has a 'physics engine' (bolstered by OAI's own claims about 'world simulation'). Like the debate about world models in LLMs, this question is both genuinely interesting and somewhat ill-defined. 🧵1/
Well, RIP nginx, long live freenginx: http://freenginx.org/pipermail/nginx/2024-February/000000.html
A thorough report of a software supply chain attack on the #PyTorch project using self-hosted GitHub runners. “Our exploit path resulted in the ability to upload malicious PyTorch releases to GitHub, upload releases to AWS, potentially add code to the main repository branch, backdoor PyTorch dependencies” https://johnstawinski.com/2024/01/11/playing-with-fire-how-we-executed-a-critical-supply-chain-attack-on-pytorch/
I wonder if the reviewer comment 'this was written by ChatGPT' is going to replace 'get a native English speaker to edit it'.
Reviewers, please, don't do either.
I get it, some papers are badly written. As a reviewer, you want to help the author convey their science, but either phrase assumes something that might not be true.
Just be factual in your review, 'I struggled to understand the conclusions because the writing was unclear.'
"Women in tech dispel a study which saw 80% of men surveyed in the industry saying there is gender parity. Here’s what women think about how things can improve."
Four in five men in tech say women are treated equally, as women criticise ‘invisible challenges’
Ever wanted to do spatial clustering of orign-destination (OD) data? Well you may want to now! University of Leeds Institute for Transport Studies (ITS) and Alan Turing Institute PhD student Hussein Mahfouz has created this early-stage visualisation of spatial clustering of these zone-zone flows 🏗️ Case study of #Leeds. Looks beautiful AND useful 🎉
Good morning! It’s the first Tuesday in February, and so you’re all invited to look through Wikipedia’s List of common misconceptions (https://en.wikipedia.org/wiki/List_of_common_misconceptions) per xkcd custom.
@arclight @JosetAEtzel @eamon @clayote Could you explain why you think language is a problem? I think it's the need to rely on large amounts of ever changing packages, which I don't think is something we can realistically change. It's all fun and nice to keep dependencies to a minimum until you can't anymore because otherwise your project won't be finished.
Learning to code in 30 lessons
My 15-year-old nephew wants to learn to program, and I will teach him.
This is the first blog post of our journey together detailing how to set up the environment we will use for the rest of the classes.
https://yabellini.netlify.app/blog/2024_learningtocode/01-learningtocode/
#rstats #100DaysToOffload - day 10
Senior lecturer at the Zhejiang-Edinburgh Joint Institute (ZJE) and Edinburgh University.
Undergraduate Programme Coordinator, Biomedical Informatics at ZJE.
I teach #imageanalysis & #dataanalysis with #RStats & #python. I study #heterogeneity in #pituitary (and other) cells.
I'm also very interested in #reproducibility and #openscience.