Robert Peharz
@ropeharz.bsky.social
440 followers 230 following 62 posts
Assistant prof at TU Graz, formerly assistant prof at TU Eindhoven, Marie-Curie Fellow at University of Cambridge. Probabilistic Machine Learning.
Posts Media Videos Starter Packs
Pinned
ropeharz.bsky.social
Amazing opportunity for #Neurosymbolic folks! 🚨🚨🚨

We are looking for a Tenure Track Prof for the 🇦🇹 #FWF Cluster of Excellence Bilateral AI (think #NeSy ++) www.bilateral-ai.net A nice starting pack for fully funded PhDs is included.

jobs.tugraz.at/en/jobs/226f...
BilateralAI – Bilateral AI – Cluster of Excellence
www.bilateral-ai.net
Reposted by Robert Peharz
ellis.eu
ELLIS @ellis.eu · Sep 4
🚀 Put your #AIStartup on the European stage!

At the #EurIPS Startup Village, get a booth + full pass, pitch your vision, connect with researchers, talent & investors, and join the inaugural showcase shaping Europe’s AI future at @euripsconf.bsky.social.

Apply by Oct 30—space is limited 👉
Startup Village - A NeurIPS-endorsed conference in Europe
Apply to EurIPS 2025 Startup Village (Dec 3–7, Copenhagen) for booth visibility, pitching, talent access, and research networking.
eurips.cc
Reposted by Robert Peharz
euripsconf.bsky.social
Is your company interested in reaching Europe's leading AI researchers?

If so, don't worry! There are still plenty of opportunity to support #EurIPS as a sponsor. Sponsorship packages are available and can be further customized if necessary.

More info at: eurips.cc/become-sponsor/
ropeharz.bsky.social
The @neuripsconf.bsky.social review system on @openreview.bsky.social has gotten perhaps sliiightly overloaded
ropeharz.bsky.social
(V)iolet (C)rumble dimension
Reposted by Robert Peharz
arnosolin.bsky.social
📣 Please share: We invite submissions to the 29th International Conference on Artificial Intelligence and Statistics (#AISTATS 2026) and welcome paper submissions at the intersection of AI, machine learning, statistics, and related areas. [1/3]
ropeharz.bsky.social
I guess it's great to know about it but, gosh, all the detail... Perhaps variable elimination is a good compromise.
ropeharz.bsky.social
I'll weigh in with Bayesian networks: excellent to understand conditional independence, probabilistic structure and information flow.
ropeharz.bsky.social
A curious @neuripsconf.bsky.social experience: review obviously LLM-generated (but with positive score), we report this to the AC. AC tells us to engage nevertheless (?) and is also poking the reviewer. The discussion just becoming more and more obscure...
ropeharz.bsky.social
Some people just like being cruel...
lookitup.baby
Nirvana is as far away from now as Elvis was from Nirvana
ropeharz.bsky.social
Many thanks for the best paper award at the #TPM workshop!
auai.org
uai2025 @auai.org · Aug 2
memories from the last day of #UAI2025:

the 8th Workshop on Tractable Probabilistic Modeling #TPM2025

tractable-probabilistic-modeling.github.io/tpm2025/
ropeharz.bsky.social
Let's say the more submissions the truer the statement 😅
About doubts of a proof and these things: yes, that's what I meant with "not substantially worse".
ropeharz.bsky.social
Absolutely. We would have #EurIPS now in Copenhagen, but, presentation there is just optional. Each paper must still be presented live in San Diego. This whole @neuripsconf.bsky.social ethical correctness yada yada is merely a fig leaf.
trappmartin.bsky.social
The world is a beautiful and fragile place that we are allowed to enjoy temporarily and should feel responsible for.

It’s regrettable that we (the ML community) don’t manage to have decentralised conferences across the globe (look at you, @neuripsconf.bsky.social).

Let's be better together!
ropeharz.bsky.social
Hot take (as AC and author): the whole #NeurIPS review process, as is, would not become substantially worse if we just left away the whole rebuttal and discussion phase.
ropeharz.bsky.social
The trick adopted from #probabilistic #circuits: make PCs deep and structured, representing an exponential number of polynomials! DeepPCE nicely scales to thousands of inputs and is competitive with standard deep neural nets. Big bonus: tractable expectations, covariances and Sobol indices!
ropeharz.bsky.social
PCE, a classical physical surrogate model, is a linear combination of (tensorproducts of) orthogonal polynomials. But, it doesn't scale gracefully to high dimensions, as the number of polynomials grows exponentially in the number of input dimensions. Also sparseness only helps to a limited extent.
ropeharz.bsky.social
Many thanks for the best paper award for our Deep Polynomial Chaos Expansion (DeepPCE), received from the #UAI workshop #tractable #probabilistic #modeling #TPM! Many congrats to the team, especially leading author Johannes Exenberger!

The workshop paper:
arxiv.org/abs/2507.21273
Reposted by Robert Peharz
ilsebyl.bsky.social
My maternity leave project is now somewhat out: I made a jupyterbook about the basics of ML, that I teach at TUE. You can check it out here:
sibylse.github.io/TUEML/intro....
The linA part is not yet formulated out and there are other todos, but maybe it helps someone with their own course design 😅
Welcome to the Data Mining and Machine Learning Book — Data Mining and Machine Learning Jupyter Book
sibylse.github.io
Reposted by Robert Peharz
ellis.eu
ELLIS @ellis.eu · Jul 16
📢 Present your NeurIPS paper in Europe!

Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date:

📅 Dec 2–7, 2025
📍 Copenhagen 🇩🇰
🔗eurips.cc

#EurIPS
@euripsconf.bsky.social
Reposted by Robert Peharz
neuripsconf.bsky.social
NeurIPS is endorsing EurIPS, an independently-organized meeting which will offer researchers an opportunity to additionally present NeurIPS work in Europe concurrently with NeurIPS.

Read more in our blog post and on the EurIPS website:
blog.neurips.cc/2025/07/16/n...
eurips.cc
eurips.cc
A NeurIPS-endorsed conference in Europe held in Copenhagen, Denmark
eurips.cc
Reposted by Robert Peharz
euripsconf.bsky.social
EurIPS is coming! 📣 Mark your calendar for Dec. 2-7, 2025 in Copenhagen 📅

EurIPS is a community-organized conference where you can present accepted NeurIPS 2025 papers, endorsed by @neuripsconf.bsky.social and @nordicair.bsky.social and is co-developed by @ellis.eu

eurips.cc
ropeharz.bsky.social
As I have PERSONALLY worked with him, I can say with certainty that it's the right guy.
ropeharz.bsky.social
Hey Prof, not sure whether I passed the exam today, since I didn't bother studying for the subject you are teaching (lol). Anyway, I realize now that I'm in trouble and really need a positive grade very soon. Can you do some extra arrangements for your favourite out of 640 students??
Reposted by Robert Peharz
nolovedeeplearning.bsky.social
#Nordic #probabilistic #AI starts in Trondheim 🇳🇴

I'll be here for a week TAing and giving a lecture on ⚡ #tractable⚡ models

and I'm in good company @andresmasegosa.bsky.social @vabor112.bsky.social @jesfrellsen.bsky.social and more covering all aspects of 🎲 AI

👉 nordic.probabilistic.ai#lecturers