Jelle Zuidema 🟥
@wzuidema.bsky.social
Associate Professor of Natural Language Processing & Explainable AI, University of Amsterdam, ILLC
Pinned
Jelle Zuidema 🟥
@wzuidema.bsky.social
· Dec 30
Transformer Interpretability 5: What is a circuit, and how does it explain LLM behavior?
YouTube video by ILLC Science
m.youtube.com
Perhaps of interest to some: we made our own series of videos on interpretability, that are visually not quite as spectacular, but add some more depth and some important context. E.g., Sandro Pezelle and Michael Hanna discuss
*how* you can find key 'circuits'
m.youtube.com/watch?v=Jfuk...
2/n
*how* you can find key 'circuits'
m.youtube.com/watch?v=Jfuk...
2/n
Reposted by Jelle Zuidema 🟥
"The man who co-discovered the double helix, perhaps not surprisingly, regarded DNA as the ultimate puppet master, immeasurably more powerful than the social and other forces that lesser (much lesser) scientists studied. Then his hubris painted him into a corner."
A Sharon Begley byline, almost 5 years after her death.
Upon hearing the news James Watson had died, a STAT reporter said in our Slack, "I wish I could read what Sharon would have written."
Incredible news: Sharon in fact did pre-write a Watson obit. And it is masterful and excoriating.
🧪🧬🧫
Upon hearing the news James Watson had died, a STAT reporter said in our Slack, "I wish I could read what Sharon would have written."
Incredible news: Sharon in fact did pre-write a Watson obit. And it is masterful and excoriating.
🧪🧬🧫
James Watson, dead at 97, was a scientific legend and a pariah among his peers
James Watson, the co-discoverer of the structure of DNA who died Thursday at 97, was a scientific legend and a pariah among his peers.
www.statnews.com
November 9, 2025 at 6:53 AM
"The man who co-discovered the double helix, perhaps not surprisingly, regarded DNA as the ultimate puppet master, immeasurably more powerful than the social and other forces that lesser (much lesser) scientists studied. Then his hubris painted him into a corner."
@mikexcohen.bsky.social Nice first post on gender bias in LLMs!
mikexcohen.substack.com/p/gender-bia...
Now nervously waiting for part 2, to see whether you'll talk your readers through our "adapting Transformer components" work. Or is there something even better?
aclanthology.org/2023.blackbo...
mikexcohen.substack.com/p/gender-bia...
Now nervously waiting for part 2, to see whether you'll talk your readers through our "adapting Transformer components" work. Or is there something even better?
aclanthology.org/2023.blackbo...
Identifying and Adapting Transformer-Components Responsible for Gender Bias in an English Language Model
Abhijith Chintam, Rahel Beloch, Willem Zuidema, Michael Hanna, Oskar van der Wal. Proceedings of the 6th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP. 2023.
aclanthology.org
November 6, 2025 at 3:13 PM
@mikexcohen.bsky.social Nice first post on gender bias in LLMs!
mikexcohen.substack.com/p/gender-bia...
Now nervously waiting for part 2, to see whether you'll talk your readers through our "adapting Transformer components" work. Or is there something even better?
aclanthology.org/2023.blackbo...
mikexcohen.substack.com/p/gender-bia...
Now nervously waiting for part 2, to see whether you'll talk your readers through our "adapting Transformer components" work. Or is there something even better?
aclanthology.org/2023.blackbo...
Dit Klavertje (dat mijn zoekmachine toont in een zoektocht naar iets heel anders), leest als een halve roman in één zin. Ik ben nu wel benieuwd naar de tweede helft, maar de link is dood (net als de ouders).
November 2, 2025 at 9:24 PM
Dit Klavertje (dat mijn zoekmachine toont in een zoektocht naar iets heel anders), leest als een halve roman in één zin. Ik ben nu wel benieuwd naar de tweede helft, maar de link is dood (net als de ouders).
Nice post on 'inoculation theory', the idea that we can fight misinformation by training/nudging people ahead of the wave of misinformation. Tom Stafford discusses the current evidence for and against this theory (cf. @profsanderlinden.bsky.social).
tomstafford.substack.com/p/helping-pe...
tomstafford.substack.com/p/helping-pe...
Helping people spot misinformation
And the greatest gift psychology gave the world
tomstafford.substack.com
November 1, 2025 at 7:59 PM
Nice post on 'inoculation theory', the idea that we can fight misinformation by training/nudging people ahead of the wave of misinformation. Tom Stafford discusses the current evidence for and against this theory (cf. @profsanderlinden.bsky.social).
tomstafford.substack.com/p/helping-pe...
tomstafford.substack.com/p/helping-pe...
Reposted by Jelle Zuidema 🟥
The Majority AI View - Anil Dash
Even though AI has been the most-talked-about topic in tech for a few years, we’re in an unusual situation where the most common opinion about AI within the tech industry is barely ever mentioned.
Even though AI has been the most-talked-about topic in tech for a few years, we’re in an unusual situation where the most common opinion about AI within the tech industry is barely ever mentioned.
The Majority AI View - Anil Dash
A blog about making culture. Since 1999.
www.anildash.com
October 18, 2025 at 8:40 PM
The Majority AI View - Anil Dash
Even though AI has been the most-talked-about topic in tech for a few years, we’re in an unusual situation where the most common opinion about AI within the tech industry is barely ever mentioned.
Even though AI has been the most-talked-about topic in tech for a few years, we’re in an unusual situation where the most common opinion about AI within the tech industry is barely ever mentioned.
Reposted by Jelle Zuidema 🟥
Twenty-four years ago today, our paper “A forkhead-domain gene is mutated in a severe speech and language disorder” was published: www.nature.com/articles/350....
A personal thread about the ups & downs of the journey we took to get to that point....1/n
🗣️🧬🧪
A personal thread about the ups & downs of the journey we took to get to that point....1/n
🗣️🧬🧪
October 4, 2025 at 1:34 PM
Twenty-four years ago today, our paper “A forkhead-domain gene is mutated in a severe speech and language disorder” was published: www.nature.com/articles/350....
A personal thread about the ups & downs of the journey we took to get to that point....1/n
🗣️🧬🧪
A personal thread about the ups & downs of the journey we took to get to that point....1/n
🗣️🧬🧪
Reposted by Jelle Zuidema 🟥
Announcing the first (and perhaps only) Multilingual Minds and Machines Meeting! Come join us in Nijmegen, June 22-23, 2026, if you are interested in computational models of human multilingualism: mmmm2026.github.io
September 19, 2025 at 11:27 AM
Announcing the first (and perhaps only) Multilingual Minds and Machines Meeting! Come join us in Nijmegen, June 22-23, 2026, if you are interested in computational models of human multilingualism: mmmm2026.github.io
Let's seek shelter under those trees over there! 😱
Fotograaf Debbie Parker legde deze "bliksem" vast in West Virginia.
August 31, 2025 at 8:27 AM
Let's seek shelter under those trees over there! 😱
This is embarrassing reporting, where AI-is-almost-sentient is the "nuanced view", and @mustafasuleymanai.bsky.social the radical.
One of @robert-booth.bsky.social's sources: Ufair "a small, undeniably fringe organisation, led by three humans and seven AIs with names such as Aether & Buzz."😱
One of @robert-booth.bsky.social's sources: Ufair "a small, undeniably fringe organisation, led by three humans and seven AIs with names such as Aether & Buzz."😱
Can AIs suffer? Big tech and users grapple with one of most unsettling questions of our times
Can AIs suffer? Big tech and users grapple with one of most unsettling questions of our times
As first AI-led rights advocacy group is founded, industry is divided on whether models are, or can be, sentient
www.theguardian.com
August 26, 2025 at 9:26 AM
This is embarrassing reporting, where AI-is-almost-sentient is the "nuanced view", and @mustafasuleymanai.bsky.social the radical.
One of @robert-booth.bsky.social's sources: Ufair "a small, undeniably fringe organisation, led by three humans and seven AIs with names such as Aether & Buzz."😱
One of @robert-booth.bsky.social's sources: Ufair "a small, undeniably fringe organisation, led by three humans and seven AIs with names such as Aether & Buzz."😱
Reposted by Jelle Zuidema 🟥
New blog post: AI Allergy.
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
AI Allergy
I remember being excited about AI. I remember 20 years ago, being excited about neuroevolutionary methods for learning adaptive behaviors in...
togelius.blogspot.com
August 13, 2025 at 5:00 AM
New blog post: AI Allergy.
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
On my increasing disgust with the AI discourse, even though I still like the technical and philosophical. And how I wish I could be excited about AI again.
togelius.blogspot.com/2025/08/ai-a...
Reposted by Jelle Zuidema 🟥
Had such a great time presenting our tutorial on Interpretability Techniques for Speech Models at #Interspeech2025! 🔍
For anyone looking for an introduction to the topic, we've now uploaded all materials to the website: interpretingdl.github.io/speech-inter...
For anyone looking for an introduction to the topic, we've now uploaded all materials to the website: interpretingdl.github.io/speech-inter...
August 19, 2025 at 9:23 PM
Had such a great time presenting our tutorial on Interpretability Techniques for Speech Models at #Interspeech2025! 🔍
For anyone looking for an introduction to the topic, we've now uploaded all materials to the website: interpretingdl.github.io/speech-inter...
For anyone looking for an introduction to the topic, we've now uploaded all materials to the website: interpretingdl.github.io/speech-inter...
Reposted by Jelle Zuidema 🟥
The heartbreaking thing about this
is that there’s already a proven way
to invest lots of money in a knowledge machine
that produces unforeseeable results
that include fantastically profitable ideas
(and some life-saving ones)
and generally benefit society
and this machine is called
a university
is that there’s already a proven way
to invest lots of money in a knowledge machine
that produces unforeseeable results
that include fantastically profitable ideas
(and some life-saving ones)
and generally benefit society
and this machine is called
a university
From the replies (bsky.app/profile/dasb...) here's Sam Altman doing what the quoted post described. He seems serious about it.
August 19, 2025 at 5:33 PM
The heartbreaking thing about this
is that there’s already a proven way
to invest lots of money in a knowledge machine
that produces unforeseeable results
that include fantastically profitable ideas
(and some life-saving ones)
and generally benefit society
and this machine is called
a university
is that there’s already a proven way
to invest lots of money in a knowledge machine
that produces unforeseeable results
that include fantastically profitable ideas
(and some life-saving ones)
and generally benefit society
and this machine is called
a university
Thought provoking work from my Amsterdam/ILLC colleagues
We built the simplest possible social media platform. No algorithms. No ads. Just LLM agents posting and following.
It still became a polarization machine.
Then we tried six interventions to fix social media.
The results were… not what we expected.
arxiv.org/abs/2508.03385
It still became a polarization machine.
Then we tried six interventions to fix social media.
The results were… not what we expected.
arxiv.org/abs/2508.03385
Can We Fix Social Media? Testing Prosocial Interventions using Generative Social Simulation
Social media platforms have been widely linked to societal harms, including rising polarization and the erosion of constructive debate. Can these problems be mitigated through prosocial interventions?...
arxiv.org
August 7, 2025 at 6:56 AM
Thought provoking work from my Amsterdam/ILLC colleagues
I'll be in Vienna only from tomorrow, but today my star PhD student Marianne is already presenting some of our work:
BLIMP-NL, in which we create a large new dataset for syntactic evaluation of Dutch LLMs, and learn a lot about dataset creation, LLM evaluation and grammatical abilities on the way.
BLIMP-NL, in which we create a large new dataset for syntactic evaluation of Dutch LLMs, and learn a lot about dataset creation, LLM evaluation and grammatical abilities on the way.
Next week I’ll be in Vienna for my first *ACL conference! 🇦🇹✨
I will present our new BLiMP-NL dataset for evaluating language models on Dutch syntactic minimal pairs and human acceptability judgments ⬇️
🗓️ Tuesday, July 29th, 16:00-17:30, Hall X4 / X5 (Austria Center Vienna)
I will present our new BLiMP-NL dataset for evaluating language models on Dutch syntactic minimal pairs and human acceptability judgments ⬇️
🗓️ Tuesday, July 29th, 16:00-17:30, Hall X4 / X5 (Austria Center Vienna)
July 29, 2025 at 9:46 AM
I'll be in Vienna only from tomorrow, but today my star PhD student Marianne is already presenting some of our work:
BLIMP-NL, in which we create a large new dataset for syntactic evaluation of Dutch LLMs, and learn a lot about dataset creation, LLM evaluation and grammatical abilities on the way.
BLIMP-NL, in which we create a large new dataset for syntactic evaluation of Dutch LLMs, and learn a lot about dataset creation, LLM evaluation and grammatical abilities on the way.
Reposted by Jelle Zuidema 🟥
Find our paper (w/ Michelle Suijkerbuijk, Zoë Prins, Jelle Zuidema @wzuidema.bsky.social, & Stefan Frank @stefanfrank.bsky.social), and dataset here:
📑 doi.org/10.1162/coli...
🗃️ doi.org/10.34973/tj4...
📑 doi.org/10.1162/coli...
🗃️ doi.org/10.34973/tj4...
BLiMP-NL: A Corpus of Dutch Minimal Pairs and Acceptability Judgments for Language Model Evaluation
Abstract. We present a corpus of 8400 Dutch sentence pairs, intended primarily for the grammatical evaluation of language models. Each pair consists of a grammatical sentence and a minimally different...
doi.org
July 24, 2025 at 3:30 PM
Find our paper (w/ Michelle Suijkerbuijk, Zoë Prins, Jelle Zuidema @wzuidema.bsky.social, & Stefan Frank @stefanfrank.bsky.social), and dataset here:
📑 doi.org/10.1162/coli...
🗃️ doi.org/10.34973/tj4...
📑 doi.org/10.1162/coli...
🗃️ doi.org/10.34973/tj4...
Reposted by Jelle Zuidema 🟥
The @interspeech.bsky.social early registration deadline is coming up in a few days!
Want to learn how to analyze the inner workings of speech processing models? 🔍 Check out the programme for our tutorial:
interpretingdl.github.io/speech-inter... & sign up through the conference registration form!
Want to learn how to analyze the inner workings of speech processing models? 🔍 Check out the programme for our tutorial:
interpretingdl.github.io/speech-inter... & sign up through the conference registration form!
Interpretability Techniques for Speech Models — Tutorial @ Interspeech 2025
interpretingdl.github.io
June 13, 2025 at 5:18 AM
The @interspeech.bsky.social early registration deadline is coming up in a few days!
Want to learn how to analyze the inner workings of speech processing models? 🔍 Check out the programme for our tutorial:
interpretingdl.github.io/speech-inter... & sign up through the conference registration form!
Want to learn how to analyze the inner workings of speech processing models? 🔍 Check out the programme for our tutorial:
interpretingdl.github.io/speech-inter... & sign up through the conference registration form!
Reposted by Jelle Zuidema 🟥
Deadlines for PhD and Postdoc vacancies coming up: applications open until Monday June 2!
Now open: vacancies for a PhD or Postdoc position to develop "Mathematical Foundations for Explainable AI" with me.
This is a new research direction that I am very excited about, and which will really start to take off over the next few years.
Come join my group: www.timvanerven.nl#open-phd-and...
This is a new research direction that I am very excited about, and which will really start to take off over the next few years.
Come join my group: www.timvanerven.nl#open-phd-and...
Tim van Erven
Tim van Erven’s website
www.timvanerven.nl
May 30, 2025 at 8:44 AM
Deadlines for PhD and Postdoc vacancies coming up: applications open until Monday June 2!
Reposted by Jelle Zuidema 🟥
This Friday at 3pm, Amsterdam's Computational Linguistics Seminar and the CLClab will host Melanie Mitchell, for a special edition of the Amsterdam Lectures in AI and Society. She will speak about "AI's Challenge of Understanding the World". Zoom link/ register:
clclab.netlify.app/2025/05/07/a...
clclab.netlify.app/2025/05/07/a...
Amsterdam Lectures in AI and Society ‹ clclab
Melanie Mitchell, leading AI researcher from the Santa Fe Institute and key voice in the discussions on the abilities and inabilities of Large Language Models, will speak at Amsterdam Science Park.
clclab.netlify.app
May 14, 2025 at 7:19 AM
This Friday at 3pm, Amsterdam's Computational Linguistics Seminar and the CLClab will host Melanie Mitchell, for a special edition of the Amsterdam Lectures in AI and Society. She will speak about "AI's Challenge of Understanding the World". Zoom link/ register:
clclab.netlify.app/2025/05/07/a...
clclab.netlify.app/2025/05/07/a...
TIL that professor Kunihiko Fukushima, inventor of the neocognitron, is still an active researcher in his late 90s. The neocognitron is a convolutional neutral network (CNN); CNNs is turn were the model family with which deep learning revolution in Artificial Intelligence started. A living legend!
May 8, 2025 at 12:33 PM
TIL that professor Kunihiko Fukushima, inventor of the neocognitron, is still an active researcher in his late 90s. The neocognitron is a convolutional neutral network (CNN); CNNs is turn were the model family with which deep learning revolution in Artificial Intelligence started. A living legend!
Reposted by Jelle Zuidema 🟥
✨New paper ✨
Introducing 🌍MultiBLiMP 1.0: A Massively Multilingual Benchmark of Minimal Pairs for Subject-Verb Agreement, covering 101 languages!
We present over 125,000 minimal pairs and evaluate 17 LLMs, finding that support is still lacking for many languages.
🧵⬇️
Introducing 🌍MultiBLiMP 1.0: A Massively Multilingual Benchmark of Minimal Pairs for Subject-Verb Agreement, covering 101 languages!
We present over 125,000 minimal pairs and evaluate 17 LLMs, finding that support is still lacking for many languages.
🧵⬇️
April 7, 2025 at 2:56 PM
✨New paper ✨
Introducing 🌍MultiBLiMP 1.0: A Massively Multilingual Benchmark of Minimal Pairs for Subject-Verb Agreement, covering 101 languages!
We present over 125,000 minimal pairs and evaluate 17 LLMs, finding that support is still lacking for many languages.
🧵⬇️
Introducing 🌍MultiBLiMP 1.0: A Massively Multilingual Benchmark of Minimal Pairs for Subject-Verb Agreement, covering 101 languages!
We present over 125,000 minimal pairs and evaluate 17 LLMs, finding that support is still lacking for many languages.
🧵⬇️
Seems like a crazy way to spend these resources. I'm all for helping American scholars move to Europe, but those that already have an ERC grant waiting for them are pretty well off. The extra 1M€ would be better used to help *other* scholars, and not to make them dependent on those lucky colleagues.
💡 The @erc.europa.eu is doubling its relocation allowance to help scientists move to Europe. Researchers from outside the EU can now receive up to €2 million in #funding to cover startup costs when taking up ERC grants in Europe.
📌 Read more:
📌 Read more:
ERC to double allowance for researchers moving to EU
The European Research Council (ERC) is doubling the relocation allowance it makes to researchers coming to the EU to take up one of its grants. At present it offers third-country researchers up to €1…
buff.ly
March 25, 2025 at 12:24 PM
Seems like a crazy way to spend these resources. I'm all for helping American scholars move to Europe, but those that already have an ERC grant waiting for them are pretty well off. The extra 1M€ would be better used to help *other* scholars, and not to make them dependent on those lucky colleagues.
Encouraging to see 5000+ protesters at our local Amsterdam protests against the budget cuts in higher education in the NL! In this time of geopolitical tensions, changing climate, democratic backsliding, & tech revolutions, governments should *invest* in knowledge & the future of the next generation
March 17, 2025 at 2:36 PM
Encouraging to see 5000+ protesters at our local Amsterdam protests against the budget cuts in higher education in the NL! In this time of geopolitical tensions, changing climate, democratic backsliding, & tech revolutions, governments should *invest* in knowledge & the future of the next generation
Reposted by Jelle Zuidema 🟥
Reposted by Jelle Zuidema 🟥
Focus on a different child each time you watch 😂
📽️SportsCenter
📽️SportsCenter
March 1, 2025 at 8:22 PM
Focus on a different child each time you watch 😂
📽️SportsCenter
📽️SportsCenter
Reposted by Jelle Zuidema 🟥
Come present your #NLP work in Leuven!
🎈 The CLIN35 Call for Abstracts is out!
clin35.ccl.kuleuven.be/call-for-abs...
The conference on Computational Linguistics in the Netherlands will take place in the ❤️ of Leuven. We're happy to host you on September 12th! 🇧🇪
Looking forward to read and hear more about your recent work at #CLIN35!
clin35.ccl.kuleuven.be/call-for-abs...
The conference on Computational Linguistics in the Netherlands will take place in the ❤️ of Leuven. We're happy to host you on September 12th! 🇧🇪
Looking forward to read and hear more about your recent work at #CLIN35!
CLIN35 - Call for Abstracts
We invite submissions for CLIN35, the 35th edition of the Computational Linguistics in the Netherlands (CLIN) conference, which will take place in Leuven on September 12th, 2025.
Abstracts describing ...
clin35.ccl.kuleuven.be
February 25, 2025 at 7:49 AM
Come present your #NLP work in Leuven!