Martin Trapp
@trappmartin.bsky.social
Assistant Prof in ML @ KTH 🇸🇪.
Previous: Aalto University 🇫🇮, TU Graz 🇦🇹, originally from 🇩🇪.
Doing: Reliable ML | uncertainty stuff | Bayesian stats | probabilistic circuits
https://trappmartin.github.io/
Previous: Aalto University 🇫🇮, TU Graz 🇦🇹, originally from 🇩🇪.
Doing: Reliable ML | uncertainty stuff | Bayesian stats | probabilistic circuits
https://trappmartin.github.io/
Reposted by Martin Trapp
Reposted by Martin Trapp
Guys, I need help! I have an empty glass wall at my office that needs to be filled.
I need memes!!!!
I need memes!!!!
November 14, 2025 at 8:42 PM
Guys, I need help! I have an empty glass wall at my office that needs to be filled.
I need memes!!!!
I need memes!!!!
Guys, I need help! I have an empty glass wall at my office that needs to be filled.
I need memes!!!!
I need memes!!!!
November 14, 2025 at 8:42 PM
Guys, I need help! I have an empty glass wall at my office that needs to be filled.
I need memes!!!!
I need memes!!!!
Reposted by Martin Trapp
Speaking of which, @ai_risks is giving away $350k in research grants for open-weight model safety. The RFP is open until December 7.
docs.google.com/forms/d/e/1...
docs.google.com/forms/d/e/1...
Open-Weight Safeguards Proposal
docs.google.com
November 12, 2025 at 2:04 PM
Speaking of which, @ai_risks is giving away $350k in research grants for open-weight model safety. The RFP is open until December 7.
docs.google.com/forms/d/e/1...
docs.google.com/forms/d/e/1...
Those repeatedly low-quality reviews are starting to get under my skin.
@iclr-conf.bsky.social, are we checking for AI-generated and low-quality reviews this year?
@iclr-conf.bsky.social, are we checking for AI-generated and low-quality reviews this year?
November 12, 2025 at 8:44 AM
Those repeatedly low-quality reviews are starting to get under my skin.
@iclr-conf.bsky.social, are we checking for AI-generated and low-quality reviews this year?
@iclr-conf.bsky.social, are we checking for AI-generated and low-quality reviews this year?
Reposted by Martin Trapp
I am hiring to my group in @ellisinstitute.fi and @aifunmcr.bsky.social, DL Dec 1
Topics: Multimodal foundation models, out-of-distribution deployable machine learning, collaborative machine learning
kaski-lab.com
Topics: Multimodal foundation models, out-of-distribution deployable machine learning, collaborative machine learning
kaski-lab.com
Samuel Kaski
kaski-lab.com
November 7, 2025 at 11:06 AM
I am hiring to my group in @ellisinstitute.fi and @aifunmcr.bsky.social, DL Dec 1
Topics: Multimodal foundation models, out-of-distribution deployable machine learning, collaborative machine learning
kaski-lab.com
Topics: Multimodal foundation models, out-of-distribution deployable machine learning, collaborative machine learning
kaski-lab.com
Reposted by Martin Trapp
@pranav-nlp.bsky.social and I are surveying researchers about naming and name changes in academia (especially computer science).
If your academic name is / has been / might someday be different from other names you've used, please tell us about it here: forms.cloud.microsoft/e/E0XXBmZdEP
If your academic name is / has been / might someday be different from other names you've used, please tell us about it here: forms.cloud.microsoft/e/E0XXBmZdEP
November 7, 2025 at 2:57 PM
@pranav-nlp.bsky.social and I are surveying researchers about naming and name changes in academia (especially computer science).
If your academic name is / has been / might someday be different from other names you've used, please tell us about it here: forms.cloud.microsoft/e/E0XXBmZdEP
If your academic name is / has been / might someday be different from other names you've used, please tell us about it here: forms.cloud.microsoft/e/E0XXBmZdEP
Reposting this while sitting on a boat in Stockholm… 👀
Newly released Stockholm Declaration recommends the following reforms to publishing:
1. Academia resumes control of publishing
2. Incentive systems to merit quality, not quantity
3. Independent fraud detection and prevention
4. Legislation and policies to protect science quality and integrity
1. Academia resumes control of publishing
2. Incentive systems to merit quality, not quantity
3. Independent fraud detection and prevention
4. Legislation and policies to protect science quality and integrity
Reformation of science publishing: the Stockholm Declaration | Royal Society Open Science
Science relies on integrity and trustworthiness. But scientists under career pressure
are lured to purchase fake publications from ‘paper mills’ that use AI-generated data,
text and image fabrication....
royalsocietypublishing.org
November 5, 2025 at 7:18 PM
Reposting this while sitting on a boat in Stockholm… 👀
Reposted by Martin Trapp
The JMLR story and operating model should be widely known in academia as a clear success story for full open access. I have friends in the humanities and pure sciences that have no clue this is even possible
November 5, 2025 at 1:16 AM
The JMLR story and operating model should be widely known in academia as a clear success story for full open access. I have friends in the humanities and pure sciences that have no clue this is even possible
Reposted by Martin Trapp
🚀 We are hiring! 🚀
🔍 Join us as a Postdoctoral Researcher (fully-funded) at the Helmholtz Institute for Human-Centered AI in Munich.
🔍 Join us as a Postdoctoral Researcher (fully-funded) at the Helmholtz Institute for Human-Centered AI in Munich.
November 3, 2025 at 10:24 AM
🚀 We are hiring! 🚀
🔍 Join us as a Postdoctoral Researcher (fully-funded) at the Helmholtz Institute for Human-Centered AI in Munich.
🔍 Join us as a Postdoctoral Researcher (fully-funded) at the Helmholtz Institute for Human-Centered AI in Munich.
Reposted by Martin Trapp
Now I'm also looking for a research software engineer to implement a pile of research results to R packages loo, posterior, bayesplot, projpred, priorsense, brms or/and Python packages ArviZ, Bambi and Kulprit. Apply by email with no specific deadline (see contact info at users.aalto.fi/~ave/)
I'm now also looking for a postdoc with strong Bayesian background and interest in developing Bayesian cross-validation theory, methods and software. Apply by email with no specific deadline (see contact information at users.aalto.fi/~ave/).
Others, please share
Others, please share
I'm looking for a doctoral student with Bayesian background to work on Bayesian workflow and cross-validation (see my publication list users.aalto.fi/~ave/publica... for my recent work) at Aalto University.
Apply through the ELLIS PhD program (dl October 31) ellis.eu/news/ellis-p...
Apply through the ELLIS PhD program (dl October 31) ellis.eu/news/ellis-p...
November 3, 2025 at 11:13 AM
Now I'm also looking for a research software engineer to implement a pile of research results to R packages loo, posterior, bayesplot, projpred, priorsense, brms or/and Python packages ArviZ, Bambi and Kulprit. Apply by email with no specific deadline (see contact info at users.aalto.fi/~ave/)
Reposted by Martin Trapp
If you're a student in need of a personal website (and if you're doing research, yes, you need a website!), I keep a list of nice examples here, most of which are reusable: www.are.na/maria-antoni...
For example, I just spotted this beautiful website by Catherine Yeh: github.com/catherinesye...
For example, I just spotted this beautiful website by Catherine Yeh: github.com/catherinesye...
November 3, 2025 at 8:11 PM
If you're a student in need of a personal website (and if you're doing research, yes, you need a website!), I keep a list of nice examples here, most of which are reusable: www.are.na/maria-antoni...
For example, I just spotted this beautiful website by Catherine Yeh: github.com/catherinesye...
For example, I just spotted this beautiful website by Catherine Yeh: github.com/catherinesye...
Amazing, congrats @avehtari.bsky.social!
Great news from the Research Council of Finland: Tuomas Hytönen, Mikko Möttönen and Aki Vehtari have been selected as Academy Professors!
Warm congratulations! 💐
www.aka.fi/en/about-the...
Warm congratulations! 💐
www.aka.fi/en/about-the...
Research Council of Finland selects 15 new Academy Professors
The Research Council of Finland (RCF) has decided on funding for 15 new Academy Professorships for the period 1 January 2026–31 December 2031. The revised Academy Professorship funding scheme provides funding to researchers who are at the international forefront of their fields and whose research strives for scientific renewal and breakthroughs.
The funding is designed to support s...
www.aka.fi
October 31, 2025 at 12:12 PM
Amazing, congrats @avehtari.bsky.social!
Does someone know of good papers discussing the variance (and variance reduction schemes) of randomised trace/diagonal estimation (e.g., Hutchinson)?
October 29, 2025 at 5:52 PM
Does someone know of good papers discussing the variance (and variance reduction schemes) of randomised trace/diagonal estimation (e.g., Hutchinson)?
Reposted by Martin Trapp
I gave a keynote at @nesyconf.org on
"Symbolic Reasoning in the Age of Large Language Models"
Check out the recording if you are curious about neurosymbolic generative AI:
www.youtube.com/watch?v=Otmi...
"Symbolic Reasoning in the Age of Large Language Models"
Check out the recording if you are curious about neurosymbolic generative AI:
www.youtube.com/watch?v=Otmi...
Guy Van den Broeck - Symbolic Reasoning in the Age of Large Language Models
YouTube video by NeSy conference
www.youtube.com
October 29, 2025 at 4:59 PM
I gave a keynote at @nesyconf.org on
"Symbolic Reasoning in the Age of Large Language Models"
Check out the recording if you are curious about neurosymbolic generative AI:
www.youtube.com/watch?v=Otmi...
"Symbolic Reasoning in the Age of Large Language Models"
Check out the recording if you are curious about neurosymbolic generative AI:
www.youtube.com/watch?v=Otmi...
Reposted by Martin Trapp
Very happy and also honoured to serve in a new way, not only thinking of our @ellisinstitute.fi , our very nice collaboration within Nordic AIR, or with Northern UK, but the whole of Europe. @ellis.eu is good for Europe!
"The recent expansion of @ellis.eu with two new institutes in @ellisinsttue.bsky.social and Finland shows that Europe has the vision, resources and drive to achieve the best AI research in the world.” Our director @samikaski.bsky.social joins the ELLIS Board, read more below! 👇
@csaalto.bsky.social
@csaalto.bsky.social
The ELLIS Society welcomes its new Board. As the primary decision-making body, it will play a vital role in shaping the future of ELLIS and advancing #AI and #MachineLearning research across Europe in a time of global change.
Read the full article: ellis.eu/news/ellis-s...
Read the full article: ellis.eu/news/ellis-s...
October 28, 2025 at 12:32 PM
Very happy and also honoured to serve in a new way, not only thinking of our @ellisinstitute.fi , our very nice collaboration within Nordic AIR, or with Northern UK, but the whole of Europe. @ellis.eu is good for Europe!
Reposted by Martin Trapp
Fisher meets Feynman! 🤝
We use score matching and a trick from quantum field theory to make a product-of-experts family both expressive and efficient for variational inference.
To appear as a spotlight @ NeurIPS 2025.
#NeurIPS2025 (link below)
We use score matching and a trick from quantum field theory to make a product-of-experts family both expressive and efficient for variational inference.
To appear as a spotlight @ NeurIPS 2025.
#NeurIPS2025 (link below)
October 27, 2025 at 12:51 PM
Fisher meets Feynman! 🤝
We use score matching and a trick from quantum field theory to make a product-of-experts family both expressive and efficient for variational inference.
To appear as a spotlight @ NeurIPS 2025.
#NeurIPS2025 (link below)
We use score matching and a trick from quantum field theory to make a product-of-experts family both expressive and efficient for variational inference.
To appear as a spotlight @ NeurIPS 2025.
#NeurIPS2025 (link below)
Reposted by Martin Trapp
Or, much more likely, they are role-playing, based on their training data.
www.theguardian.com/technology/2...
www.theguardian.com/technology/2...
AI models may be developing their own ‘survival drive’, researchers say
Like 2001: A Space Odyssey’s HAL 9000, some AIs seem to resist being turned off and will even sabotage shutdown
www.theguardian.com
October 25, 2025 at 5:14 PM
Or, much more likely, they are role-playing, based on their training data.
www.theguardian.com/technology/2...
www.theguardian.com/technology/2...
Reposted by Martin Trapp
What if we did a single run and declared victory
October 23, 2025 at 2:28 AM
What if we did a single run and declared victory
Reposted by Martin Trapp
What if your CBM actually doesn’t look at attributes when it explains the classification decision? We suggest to use this dataset and method to generate realistic images with unusual attributes to stress-test explainability methods. @jessica-bader.bsky.social is at ICCV to talk to you about it!
1/
SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutions
@jessica-bader.bsky.social , @lgirrbach.bsky.social , Stephan Alaniz, @zeynepakata.bsky.social
[Paper]: arxiv.org/pdf/2507.23784
[Code]: github.com/ExplainableM...
📍Hall I #2142
🕑Oct 23, Poster Session 5
SUB: Benchmarking CBM Generalization via Synthetic Attribute Substitutions
@jessica-bader.bsky.social , @lgirrbach.bsky.social , Stephan Alaniz, @zeynepakata.bsky.social
[Paper]: arxiv.org/pdf/2507.23784
[Code]: github.com/ExplainableM...
📍Hall I #2142
🕑Oct 23, Poster Session 5
October 20, 2025 at 2:14 PM
What if your CBM actually doesn’t look at attributes when it explains the classification decision? We suggest to use this dataset and method to generate realistic images with unusual attributes to stress-test explainability methods. @jessica-bader.bsky.social is at ICCV to talk to you about it!
Reposted by Martin Trapp
LUMI AI Factory Hub is #hiring a Coordinator to run operations, coordinate activities, plan events, build community, and connect researchers, industry and students to the @lumi-supercomputer.eu and @csc.fi infrastructure via the Hub at @aalto.fi.
Apply by 19.11.: www.aalto.fi/en/open-posi...
Apply by 19.11.: www.aalto.fi/en/open-posi...
Coordinator – LUMI AI Factory Hub | Aalto University
We’re looking for a Coordinator to help shape and run the Otaniemi AI Factory Hub – a collaborative environment connecting researchers, students, and companies to accelerate breakthroughs in artificial intelligence.
www.aalto.fi
October 20, 2025 at 10:49 AM
LUMI AI Factory Hub is #hiring a Coordinator to run operations, coordinate activities, plan events, build community, and connect researchers, industry and students to the @lumi-supercomputer.eu and @csc.fi infrastructure via the Hub at @aalto.fi.
Apply by 19.11.: www.aalto.fi/en/open-posi...
Apply by 19.11.: www.aalto.fi/en/open-posi...
Reposted by Martin Trapp
Call for papers: sites.google.com/view/amortiz...
How can we make probabilistic ML fast, generalizable, and ready for the real world? 🤔
Join us for the ELLIS UnConference Workshop on Amortized ProbML on Dec 2 in Copenhagen 🇩🇰 co-located with hashtag#EurIPS
How can we make probabilistic ML fast, generalizable, and ready for the real world? 🤔
Join us for the ELLIS UnConference Workshop on Amortized ProbML on Dec 2 in Copenhagen 🇩🇰 co-located with hashtag#EurIPS
Amortized ProbML 2025 - Call for Papers
Accepted papers will be presented as posters during the workshop poster session on 2nd December, 2025.
Please be aware that this is not the same as the ELLIS UnConference poster session.
sites.google.com
October 17, 2025 at 11:22 AM
Call for papers: sites.google.com/view/amortiz...
How can we make probabilistic ML fast, generalizable, and ready for the real world? 🤔
Join us for the ELLIS UnConference Workshop on Amortized ProbML on Dec 2 in Copenhagen 🇩🇰 co-located with hashtag#EurIPS
How can we make probabilistic ML fast, generalizable, and ready for the real world? 🤔
Join us for the ELLIS UnConference Workshop on Amortized ProbML on Dec 2 in Copenhagen 🇩🇰 co-located with hashtag#EurIPS
Reposted by Martin Trapp
The abstract submission deadline for the ELLIS Workshop on Amortized Probabilistic ML has been extended to October 25th!
You still have time to submit a short abstract on your work in: Amortized Inference, Neural processes, Prior-fitted networks, Sequential Decision-Making, and Foundation models.
You still have time to submit a short abstract on your work in: Amortized Inference, Neural processes, Prior-fitted networks, Sequential Decision-Making, and Foundation models.
October 17, 2025 at 11:22 AM
The abstract submission deadline for the ELLIS Workshop on Amortized Probabilistic ML has been extended to October 25th!
You still have time to submit a short abstract on your work in: Amortized Inference, Neural processes, Prior-fitted networks, Sequential Decision-Making, and Foundation models.
You still have time to submit a short abstract on your work in: Amortized Inference, Neural processes, Prior-fitted networks, Sequential Decision-Making, and Foundation models.
I think OpenReview is really not a good system for peer review. It's mostly better than CMT and clearly better than EasyChair, but it still sucks in so many aspects (overly complex, no reviewer chat, etc).
Not sure if this is a controversial opinion or not... I fear it is not.
Not sure if this is a controversial opinion or not... I fear it is not.
October 13, 2025 at 9:13 AM
I think OpenReview is really not a good system for peer review. It's mostly better than CMT and clearly better than EasyChair, but it still sucks in so many aspects (overly complex, no reviewer chat, etc).
Not sure if this is a controversial opinion or not... I fear it is not.
Not sure if this is a controversial opinion or not... I fear it is not.
Whenever I try Le Chat by Mistral, I am surprised how much faster it is than the other system.
October 12, 2025 at 4:33 PM
Whenever I try Le Chat by Mistral, I am surprised how much faster it is than the other system.