Aniello De Santo
@anids.bsky.social
The mystery of the Universe is its comprehensibility. Computational psycholinguist, I guess, but not a professional account. Opinions etc etc my own. he\him (pfp cc. Dustin Nguyen)
I am told Outlook is also down so...freedom?
Canvas down, adding or dropping students from random courses. Amazing. 10/10 software. Worth the $$$$ the University pays for it and the tears of the many of us who hate it even when it "works".
January 22, 2026 at 10:44 PM
I am told Outlook is also down so...freedom?
Canvas down, adding or dropping students from random courses. Amazing. 10/10 software. Worth the $$$$ the University pays for it and the tears of the many of us who hate it even when it "works".
January 22, 2026 at 9:33 PM
Canvas down, adding or dropping students from random courses. Amazing. 10/10 software. Worth the $$$$ the University pays for it and the tears of the many of us who hate it even when it "works".
This is literally the plot of "Common Side Effects"
With most psychedelic drugs, you never know what you're going to get. But this mysterious mushroom from China - without fail - causes users to hallucinate tiny people: crawling up walls, popping out from under furniture and marching under doors. www.bbc.com/future/artic...
'They saw them on their dishes when eating': The mushroom making people hallucinate dozens of tiny humans
Only recently described by science, the mysterious mushrooms are found in different parts of the world, but they give people the same exact visions.
www.bbc.com
January 22, 2026 at 9:30 PM
This is literally the plot of "Common Side Effects"
Reposted by Aniello De Santo
I don't have trouble saying that a human researcher who wants to do interesting work and has minimal self-respect, when left to their own devices, will not make up citations or blend citations together (or, not do this unknowingly). For mistakes made, it will likely not be from lack of checking.
January 22, 2026 at 8:00 PM
I don't have trouble saying that a human researcher who wants to do interesting work and has minimal self-respect, when left to their own devices, will not make up citations or blend citations together (or, not do this unknowingly). For mistakes made, it will likely not be from lack of checking.
Ohhhhhhhhhhhhhh
The program for GURT 2026 ("Futuring Linguistics") is live! Gonna be a good one
docs.google.com/document/d/1...
docs.google.com/document/d/1...
GURT 2026 Schedule at a glance
Schedule at a glance Friday, March 6, 2026 10:30 – 3:00 Registration | ICC Galleria 11:00 – 12:00 Pre-conference welcome & workshop for students | ICC 12:00 – 1:30 Lunch b...
docs.google.com
January 21, 2026 at 12:21 AM
Ohhhhhhhhhhhhhh
Reposted by Aniello De Santo
Much more controversially: I think connectionists in the 80s may have erred if you asked them: would a model trained on the intellectual history of the species be a viable model of human cognition?
January 19, 2026 at 8:59 PM
Much more controversially: I think connectionists in the 80s may have erred if you asked them: would a model trained on the intellectual history of the species be a viable model of human cognition?
And it was great!
It’s incredible that they made a real movie out of Dungeons & Dragons
January 18, 2026 at 8:37 PM
And it was great!
Reposted by Aniello De Santo
This is awesome! I have been curating a list of #linguistics journals, with information on their publishing model (and also on whether they admit submissions created by LaTeX, a high-quality, professional Open Source way to produce scientific documents).
uni-duesseldorf.sciebo.de/s/BvRY0u3rDY...
uni-duesseldorf.sciebo.de/s/BvRY0u3rDY...
January 16, 2026 at 9:05 AM
This is awesome! I have been curating a list of #linguistics journals, with information on their publishing model (and also on whether they admit submissions created by LaTeX, a high-quality, professional Open Source way to produce scientific documents).
uni-duesseldorf.sciebo.de/s/BvRY0u3rDY...
uni-duesseldorf.sciebo.de/s/BvRY0u3rDY...
Woke up remembering I have no breakfast food left. Tragedy. 😭😭
January 15, 2026 at 3:12 PM
Woke up remembering I have no breakfast food left. Tragedy. 😭😭
Wish I was at that workshop
On a train early this morning to head to Cambridge for #OCP23
There's a satellite workshop on the Role of Representation in Computational Phonology today and then the main conference begins in earnest tomorrow
🐦🐦
There's a satellite workshop on the Role of Representation in Computational Phonology today and then the main conference begins in earnest tomorrow
🐦🐦
January 13, 2026 at 6:51 PM
Wish I was at that workshop
Reposted by Aniello De Santo
If you're interested in a short behind-the-scenes of the retraction of a mega-metascience paper—the first of its kind (the retraction, not the paper)—this might be for you! Had to gloss over so many details to observe allotted time but happy to share an extended version of my slide deck with links.
📽️ “Claims about scientific rigour require rigour” / with Berna Devezer @devezer.bsky.social
youtu.be/qNJ8xqaObR8?...
The event was organised by ReproducibiliTea UniBasel
youtu.be/qNJ8xqaObR8?...
The event was organised by ReproducibiliTea UniBasel
Open Science Basel Year 6 #5 - Berna Devezer
YouTube video by BAMM
youtu.be
January 13, 2026 at 3:26 PM
If you're interested in a short behind-the-scenes of the retraction of a mega-metascience paper—the first of its kind (the retraction, not the paper)—this might be for you! Had to gloss over so many details to observe allotted time but happy to share an extended version of my slide deck with links.
Love SLC’s specific brand of “free send cold, but also so polluted that the fog is smog and your allergies get triggered” 🙃
January 13, 2026 at 2:58 PM
Love SLC’s specific brand of “free send cold, but also so polluted that the fog is smog and your allergies get triggered” 🙃
Some of you out there really thinking that having an LLM estimate norms of things like “animacy” has the same validity as running an experiment asking actual humans to do it. The field is doomed.
January 11, 2026 at 11:52 PM
Some of you out there really thinking that having an LLM estimate norms of things like “animacy” has the same validity as running an experiment asking actual humans to do it. The field is doomed.
Italy just voted to end mail-ballots for Italians residing abroad so I guess now I am tied to TWO countries I have no representation power in 🙃
January 10, 2026 at 2:17 AM
Italy just voted to end mail-ballots for Italians residing abroad so I guess now I am tied to TWO countries I have no representation power in 🙃
If as a reviewer you write a sentence like “ I question why the authors did X” it would be nice if you could follow it up with what exactly you question about it, especially if X was justified by like 10 references and 2 full paragraphs of rationale 🙃
January 9, 2026 at 9:16 PM
If as a reviewer you write a sentence like “ I question why the authors did X” it would be nice if you could follow it up with what exactly you question about it, especially if X was justified by like 10 references and 2 full paragraphs of rationale 🙃
After a week of classes my students already understand that I am the kind of prof who appreciates references to Avatar The last Airbender in their assignments. success.
January 8, 2026 at 11:29 PM
After a week of classes my students already understand that I am the kind of prof who appreciates references to Avatar The last Airbender in their assignments. success.
Reposted by Aniello De Santo
Logging off to lie in bed dissociating for a while and I invite you all to join me
January 8, 2026 at 12:28 AM
Logging off to lie in bed dissociating for a while and I invite you all to join me
Was a bit demotivated today but I got to chat about some cool sentence processing results with a friend and I got excited again
January 6, 2026 at 10:39 PM
Was a bit demotivated today but I got to chat about some cool sentence processing results with a friend and I got excited again
🙃🙃🙃🙃🙃
help, my university has released new guidance on LLM use for students and i don't want to have yet another reason to YELL AT THEM
January 5, 2026 at 4:54 PM
🙃🙃🙃🙃🙃
Reposted by Aniello De Santo
My talk on the limitations of algorithmic fairness is now online! youtu.be/YjZ4s3nqxbs?...
I had the privilege of delivering my first talk at @princetoncitp.bsky.social today.
January 4, 2026 at 9:13 PM
My talk on the limitations of algorithmic fairness is now online! youtu.be/YjZ4s3nqxbs?...
One canvas page up, one more to go
January 3, 2026 at 5:34 PM
One canvas page up, one more to go
Woke up early thinking I’d be productive (semester starts Monday 😭😭) and then <waves around>
January 3, 2026 at 2:38 PM
Woke up early thinking I’d be productive (semester starts Monday 😭😭) and then <waves around>
6am flights always seem like a good idea until you have to wake up for them
December 31, 2025 at 2:04 AM
6am flights always seem like a good idea until you have to wake up for them
🔥 well-thought, theoretically grounded models >>>>>>> trendy uninterpretable forest killers
We wrote a thing -- showing you don't need LLMs to model language production dynamics like the tendency for speakers to reduce predictable words. All you have to do is better model how speech rate varies depending on where a word is and how long the utterance is. arxiv.org/abs/2512.23659
Less is more: Probabilistic reduction is best explained by small-scale predictability measures
The primary research questions of this paper center on defining the amount of context that is necessary and/or appropriate when investigating the relationship between language model probabilities and ...
arxiv.org
December 30, 2025 at 2:09 PM
🔥 well-thought, theoretically grounded models >>>>>>> trendy uninterpretable forest killers
I love caldarroste, deeply missed in cold SLC
December 29, 2025 at 8:55 PM
I love caldarroste, deeply missed in cold SLC