Aaron Steven White
@aaronstevenwhite.io
computational semanticist. into modular synths and cocktails. http://aaronstevenwhite.io
Come join us at Rochester to work on low-resource NLP with @cmdowney.bsky.social! 👇
I'm recruiting Linguistics PhD students this cycle! You should apply to work with me if you're interested in anything low-resource NLP. Some topics I'm particularly excited about currently:
November 11, 2025 at 2:20 PM
Come join us at Rochester to work on low-resource NLP with @cmdowney.bsky.social! 👇
Reposted by Aaron Steven White
lots of invigorating discussion this year at #BUCLD2025. Unsurprisingly, LLMs as models & as tools made many appearances w/a wide range of views/claims/caveats ab what they can(not) tell us ab language development
I'm still chewing on a few things that seemed to get short shrift 1/4 🐦🐦
I'm still chewing on a few things that seemed to get short shrift 1/4 🐦🐦
November 10, 2025 at 4:30 PM
lots of invigorating discussion this year at #BUCLD2025. Unsurprisingly, LLMs as models & as tools made many appearances w/a wide range of views/claims/caveats ab what they can(not) tell us ab language development
I'm still chewing on a few things that seemed to get short shrift 1/4 🐦🐦
I'm still chewing on a few things that seemed to get short shrift 1/4 🐦🐦
Undetached bus(es) parts.
November 1, 2025 at 12:16 AM
Undetached bus(es) parts.
Reposted by Aaron Steven White
I would rather live next door to someone who speaks another language than live next door to someone who doesn’t want to live near people who speak other languages. I know nothing about Person A but I know person B is a shitbird.
Vance says it is "totally reasonable and acceptable" for people to not want to live next door to people who speak a different language than they do
October 29, 2025 at 5:01 PM
I would rather live next door to someone who speaks another language than live next door to someone who doesn’t want to live near people who speak other languages. I know nothing about Person A but I know person B is a shitbird.
This language model is very 2018.
October 29, 2025 at 1:34 AM
This language model is very 2018.
Every music genre is ambient if you just vibe hard enough.
October 23, 2025 at 6:06 AM
Every music genre is ambient if you just vibe hard enough.
Reposted by Aaron Steven White
Cool project as part of an even cooler Collaborative Research Center, focused on "Common Ground"
What even is it? (Many takes abound)
Can there be a cohesive version which does what it has variously used for?
Can we square it with modern psycholing/cog psych theory?
Come help us figure out!
What even is it? (Many takes abound)
Can there be a cohesive version which does what it has variously used for?
Can we square it with modern psycholing/cog psych theory?
Come help us figure out!
Jobs: Cognitive Science, General Linguistics, Philosophy of Language, Pragmatics, Semantics: Postdoctoral Researcher (3.5 years) Philosophy of Language / Pragmatics / Logic / Cognitive Science, University of Tübingen
Jobs: Cognitive Science, General Linguistics, Philosophy of Language, Pragmatics, Semantics: Postdoctoral Researcher (3.5 years) Philosophy of Language / Pragmatics / Logic / Cognitive Science, University of Tübingen
Description:
We are looking for a highly motivated postdoctoral researcher to support a large-scale interdisciplinary research initiative on the topic of "Common Ground" at the University of Tübingen. The project seeks to investigate the notion of common ground in linguistic communication by taking an interdisciplinary perspective, incorporating insights from, e.g., the philosophy of language, the cognitive language sciences, linguistic pragmatics, and/or formal logic or epistemology. The ide
dlvr.it
October 20, 2025 at 11:11 AM
Cool project as part of an even cooler Collaborative Research Center, focused on "Common Ground"
What even is it? (Many takes abound)
Can there be a cohesive version which does what it has variously used for?
Can we square it with modern psycholing/cog psych theory?
Come help us figure out!
What even is it? (Many takes abound)
Can there be a cohesive version which does what it has variously used for?
Can we square it with modern psycholing/cog psych theory?
Come help us figure out!
This feels like it should be so deeply obvious, and I don’t understand how an entire class of institutions obsessed with signaling prestige have not caught onto the fact that, e.g., one major way people now signal wealth is by their ability to consume and collect physical media.
In about three years the entire university pivot to AI curricula and schools and programs is going to be so deeply embarrassing. We will all pretend it never happened and I will be standing there, looking at people with a mirror in my eyes. This is all so embarrassing.
October 17, 2025 at 12:53 PM
This feels like it should be so deeply obvious, and I don’t understand how an entire class of institutions obsessed with signaling prestige have not caught onto the fact that, e.g., one major way people now signal wealth is by their ability to consume and collect physical media.
Reposted by Aaron Steven White
The palpable relief of reading a syntactically awkward student paper nevertheless brimming with ideas they’re trying on after having to consume the stale cardboard generalities of a Chat GPT essay.
October 11, 2025 at 4:20 PM
The palpable relief of reading a syntactically awkward student paper nevertheless brimming with ideas they’re trying on after having to consume the stale cardboard generalities of a Chat GPT essay.
Reposted by Aaron Steven White
A company that believed it was in the verge of AGI or ASI wouldn’t capitulate to the government because it wouldn’t care about government contracts. They would soon BE the economy and the government would soon be capitulating to them.
The administration “is currently pressuring OpenAI and other AI companies to make their models more conservative-friendly.”
OpenAI is trying to clamp down on ‘bias’ in ChatGPT
GPT-5 is better at resisting liberal ‘pressure,’ the company says.
www.theverge.com
October 11, 2025 at 8:57 PM
A company that believed it was in the verge of AGI or ASI wouldn’t capitulate to the government because it wouldn’t care about government contracts. They would soon BE the economy and the government would soon be capitulating to them.
I’m going to be in Oxford Sunday afternoon through Tuesday morning of next week, London on Sunday morning and Tuesday night. DM me if you’re in either place and interested in grabbing a coffee or cocktail.
October 9, 2025 at 1:51 AM
I’m going to be in Oxford Sunday afternoon through Tuesday morning of next week, London on Sunday morning and Tuesday night. DM me if you’re in either place and interested in grabbing a coffee or cocktail.
Doing this today. Going to be so sick. I sort of miss writing bespoke samplers by hand.
Some things I'm looking forward to:
1. Building an MH sampler from the ground up in the statistical inference module and using it to characterize uncertainty about the distribution of pronominal case in different corpora.
1. Building an MH sampler from the ground up in the statistical inference module and using it to characterize uncertainty about the distribution of pronominal case in different corpora.
October 8, 2025 at 3:55 PM
Doing this today. Going to be so sick. I sort of miss writing bespoke samplers by hand.
The only Halloween movie you need.
October 7, 2025 at 12:49 PM
The only Halloween movie you need.
Reposted by Aaron Steven White
Computational Psycholinguistics Meeting 2025
cpl2025.sites.uu.nl
When: December 18–19, 2025
Where: Utrecht, the Netherlands
Abstract submission deadline: June 15, 2025
Organizers: Jakub Dotlačil, Lena Jäger, Bruno Nicenboim, Ece Takmaz
cpl2025.sites.uu.nl
When: December 18–19, 2025
Where: Utrecht, the Netherlands
Abstract submission deadline: June 15, 2025
Organizers: Jakub Dotlačil, Lena Jäger, Bruno Nicenboim, Ece Takmaz
Computational Psycholinguistics Meeting 2025 | Universiteit Utrecht
Universiteit Utrecht
cpl2025.sites.uu.nl
March 20, 2025 at 7:25 PM
Computational Psycholinguistics Meeting 2025
cpl2025.sites.uu.nl
When: December 18–19, 2025
Where: Utrecht, the Netherlands
Abstract submission deadline: June 15, 2025
Organizers: Jakub Dotlačil, Lena Jäger, Bruno Nicenboim, Ece Takmaz
cpl2025.sites.uu.nl
When: December 18–19, 2025
Where: Utrecht, the Netherlands
Abstract submission deadline: June 15, 2025
Organizers: Jakub Dotlačil, Lena Jäger, Bruno Nicenboim, Ece Takmaz
Reposted by Aaron Steven White
SALT 36 will be held at my alma mater, the University of Buenos Aires, on July 29-31 2026. This will be the first time the conference takes place in South America.
Abstract deadline: Dec 15, 2025
Link: saltconf.github.io/salt36/
Abstract deadline: Dec 15, 2025
Link: saltconf.github.io/salt36/
October 2, 2025 at 9:21 AM
SALT 36 will be held at my alma mater, the University of Buenos Aires, on July 29-31 2026. This will be the first time the conference takes place in South America.
Abstract deadline: Dec 15, 2025
Link: saltconf.github.io/salt36/
Abstract deadline: Dec 15, 2025
Link: saltconf.github.io/salt36/
Reposted by Aaron Steven White
Excited to introduce Vibes from Meta. Eat your slop, piggies!
September 30, 2025 at 5:59 PM
Excited to introduce Vibes from Meta. Eat your slop, piggies!
I've found it kind of a pain to work with resources like VerbNet, FrameNet, PropBank (frame files), and WordNet using existing tools. Maybe you have too. Here's a little package that handles data management, loading, and cross-referencing via either a CLI or a python API.
GitHub - aaronstevenwhite/glazing: Unified data models and interfaces for syntactic and semantic frame ontologies.
Unified data models and interfaces for syntactic and semantic frame ontologies. - aaronstevenwhite/glazing
github.com
September 27, 2025 at 1:51 PM
I've found it kind of a pain to work with resources like VerbNet, FrameNet, PropBank (frame files), and WordNet using existing tools. Maybe you have too. Here's a little package that handles data management, loading, and cross-referencing via either a CLI or a python API.
This was never not going to be a banger.
Michelangelo Dying, by Cate Le Bon
10 track album
catelebon.bandcamp.com
September 27, 2025 at 1:08 PM
This was never not going to be a banger.
Reposted by Aaron Steven White
🚨The text has several code snippets, which are taken from the Haskell project under active development here:
github.com/probabilisti...
github.com/probabilisti...
September 26, 2025 at 7:56 PM
🚨The text has several code snippets, which are taken from the Haskell project under active development here:
github.com/probabilisti...
github.com/probabilisti...
Reposted by Aaron Steven White
Brand new version of this paper (now a short book!) available at lingbuzz.net/lingbuzz/008...!
September 26, 2025 at 6:12 PM
Brand new version of this paper (now a short book!) available at lingbuzz.net/lingbuzz/008...!
Nothing quite like Cocteau Twins’ Blue Bell Knoll on a foggy fall evening.
September 25, 2025 at 11:03 PM
Nothing quite like Cocteau Twins’ Blue Bell Knoll on a foggy fall evening.
Reposted by Aaron Steven White
Please stop using OpenTorment!!! It's only open-core! TormentLibre is the fully FOSS alternative!
September 23, 2025 at 2:13 AM
Please stop using OpenTorment!!! It's only open-core! TormentLibre is the fully FOSS alternative!
Reposted by Aaron Steven White
I just want you guys to know that if your #rstats models aren't converging, it's likely because you're taking Tylenol.
Tylenol can flatten likelihood surfaces due to multidimensional recontainerization of typhoid surfaces.
I'm going to ask the FDA to modify the label of Tylenol to indicate this.
Tylenol can flatten likelihood surfaces due to multidimensional recontainerization of typhoid surfaces.
I'm going to ask the FDA to modify the label of Tylenol to indicate this.
September 23, 2025 at 1:30 AM
I just want you guys to know that if your #rstats models aren't converging, it's likely because you're taking Tylenol.
Tylenol can flatten likelihood surfaces due to multidimensional recontainerization of typhoid surfaces.
I'm going to ask the FDA to modify the label of Tylenol to indicate this.
Tylenol can flatten likelihood surfaces due to multidimensional recontainerization of typhoid surfaces.
I'm going to ask the FDA to modify the label of Tylenol to indicate this.
The only way to send summer off is with a zombie in the yard.
September 21, 2025 at 10:05 PM
The only way to send summer off is with a zombie in the yard.