One of my pet peeves: People are obsessed with seeing/plotting error bars while also being confused about what they should represent. blog.genesmindsmachines.com/p/the-error-...

The error bar cargo cult

You don't need to understand error bars. You just need to draw them.

blog.genesmindsmachines.com

new blog post: www.utkuturk.com/posts/bayesi... ! i want both tables and the coefficient whisker plots, but i dont like how many things just cut through the text. so, i wrote an automated script that makes nice tables with whisker plots in tikz.

While waiting for the operation, I was implementing a bayesian bimodality test in R following this paper: jammalam.faculty.pstat.ucsb.edu/html/Some%20... So hopefully will share it as a blogpost soon

Last month I got diagnosed with thyroid cancer, went through the operation 2 weeks ago, will go through some more procedures. But I am doing fine, I am healthy and I am recovering very well.

Here's my attempt at writing a Semantics Typesetting Guide for Typst: typst.app/project/rDI6... Hopefully soon I am going to write the Syntax version as well as the blog post. The best thing about Typst is that it immediately gives you a preview, compliation time is practically 0.

Turkish 🇹🇷 native speakers needed! I'm running a pilot sentence comprehension study in Turkish with an Erasmus student from Boğaziçi, it's only a short one (10 min tops I promise!). Please share and take part! Teşekkürler! farm.pcibex.net/p/eESOzY/

just writing a code that takes any pdf and if there is any bar plots, it makes it a pie chart. because no one should be using bar plots.

I have to say, none of the ordinal data plots I have seen online really make sense.

migrated a note about plotting probability allocations from ordinal regression models www.tjmahr.com/notes/2022-0...

😀 Enter the amortized Bayesian workflow from here 👉 pipme.github.io/amortized-Ba... Huge thanks to my great collaborators: @avehtari.bsky.social, @paulbuerkner.com, @stefanradev.bsky.social, @lacerbi.bsky.social, @marvin-schmitt.com!

Amortized Bayesian Workflow

Adaptive Bayesian workflow combining amortized inference with PSIS and MCMC diagnostics.

pipme.github.io

Luigi Acerbi@lacerbi.bsky.social · 3mo ago

1/ Wouldn't it be nice if you could perform Bayesian inference *efficiently* but also *reliably*? Amortized inference offers the former, while MCMC is often presented as the "gold standard" for accuracy and reliability. Enter the Amortized Bayesian Workflow...

new blog post about, like. why i teach syntax, i guess. may be a little bit inside baseball, curious about your thoughts kconrod.medium.com/build-me-a-b...

build me a bird

or, an apologia for (still) teaching syntax

kconrod.medium.com

people think that I am only harsh on their topic and talk. here's my take on experimentation, and the people's misleading trust in them. I feel like alchemy is a good allegory and a reminder for all of us to think. www.utkuturk.com/posts/alchemy/

My sunday morning: I have been using β when reporting coefficients, but I have to say I really want to go back to θ.

And the reason I like this paper so much these days is this graph that I have been thinking about. Subjs who make agreement errors in Exp3 are faster to start speaking while in Exp4 they are slower. Funnily enough this is only the case with unaccusatives.

utku turk@utkuturk.com · 4mo ago

My favorite paper these days: www.sciencedirect.com/science/arti... "[I]t is hard to predict ex ante (i.e., before data collection) when errors will be fast or slow. [...] [We] validate a [...] model which successfully predicts when errors will be faster or slower than correct responses."

My favorite paper these days: www.sciencedirect.com/science/arti... "[I]t is hard to predict ex ante (i.e., before data collection) when errors will be fast or slow. [...] [We] validate a [...] model which successfully predicts when errors will be faster or slower than correct responses."

Errors, fast and slow

Human errors in cognitive, attentional, and decision-making tasks are sometimes faster than correct responses, and sometimes slower, even for the same…

sciencedirect.com

Reducing number of figures in a paper from 48 to 22 👍(i am crying inside)

My latest cartoon for @newscientist.com

Is anyone using Typst? I realized it is a lot more easier to create documents and faster to compile with. Following what Mal did with SynTex, I want to create something with Typst, here's a lousy draft for semantics (github.com/utkuturk/typ...). Let me know what you would like to see more.

most of the things on stats i learned was thanks to first with my MA advisor Pavel, and then advanced bayesian course in SMLP. I wish the world was a better place and travel was easier. but please apply, it is awesome!

Shravan Vasishth@shravanvasishth.bsky.social · 4mo ago

Tomorrow is the last day to sign up for the tenth summer school on statistical methods for linguistics and psychology (Aug 24-28, 2026, in Potsdam, Germany): vasishth.github.io/smlp2026/

I gave a talk at HSP 2026 @mit.edu, on agreement attraction in Turkish! We showed evidence for the role of structural positional associations in memory. This was joint work with @utkuturk.com, Duygu Demiray and @linguistbrian.bsky.social (that's me at the very back!)

Will be in Boston until the Sunday for HSP, would be happy to meet with folks that is around!

Not sure what "entirely generated" means, exactly, but I feel we need new norms around this. Here's a blog I like about "etiquette", "consent", and what should be considered "rude" when sharing AI-generated content. Disclosures after the fact do not cut it, IMO. distantprovince.by/posts/its-ru...

It's rude to show AI output to people | Alex Martsinovich

Feeding slop is an act of war

distantprovince.by

New preprint! 🚨 Does surface overlap affect dependency resolution? Using evidence from Turkish agreement, I show the answer is no. I argue that "phonological modulation" of memory is better explained by a more parsimonious mechanism: statistical controllerhood association. osf.io/preprints/ps...

Illusion alert 🚨 We looked at NPI and NCI illusions in 🇨🇿. We aimed to distinguish between them by using negation. In English, it can't cause NPI illusions, but with NCIs it might, because those are licensed in the syntax. Surprisingly, we found strong illusions with both! 1/2 osf.io/preprints/ps...

www.utkuturk.com/posts/audio-... I made the mistake of recording my own voice for an experiment. Luckily, no one heard it because it was really easy (and more pleasant for participants to listen) to use text-to-speech models. with some tricks to make it more realistic (by varying some values)

Have been working (with Aron Hirsch) on two interesting cases for alternatives in semantics, that was taken for granted until very recently: polar questions. Turkish present interesting case, because the question morpheme can attach to almost any constituent and also can form canonical forms.

I do not understand why people are so mean to her, I love intermediate objects too, and I love pipes in R too, but of course you do not use neither of those all the time.

Julia M. Rohrer@dingdingpeng.the100.ci · 5mo ago

x1 <- what() x2 <- is(x1) x3 <- wrong(x2) x4 <- with(x3) you(x4, “?“)

May I suggest: github.com/MilesMcBain/...

GitHub - MilesMcBain/breakerofchains: Break your chain at the cursor line. Run the first bit. See the output. Be free.

Break your chain at the cursor line. Run the first bit. See the output. Be free. - MilesMcBain/breakerofchains

github.com

Have you ever wondered when Lunar New Year - Ramadan - Ash Wednesday is going to be on the same day? It is 3571. And after next year, the next time they are going to be in the same week is 2127. Here's the code for it: colab.research.google.com/drive/1olp7f...

Google Colab

colab.research.google.com

I love creating wikis or step-by-step guide for every procedure because I get to roleplay as that guy in the opening scene of a post-apocalyptic movie who just records everything.