iiinova — where your digital self keeps evolving

Lukas Heinrich

@lukasheinrich.com

1.1K followers 934 following 75 posts

High Energy Physics/Machine Learning/Data Science Prof @tum.de www.lukasheinrich.com

Lukas Heinrich @lukasheinrich.com · 19/05/2026

You know your PhD students are ready to graduate when they start organizing their own workshops - here is our own Malin Horstmann kicking off the SBI in @atlasexperiment.bsky.social workshop!

072

Lukas Heinrich @lukasheinrich.com · 13/05/2026

Great to be back in Pittsburgh - first time I’m visiting CMU

030

Lukas Heinrich @lukasheinrich.com · 08/05/2026

but with the benefit of tractable likelihoods, differentiability (and perhaps acceleration). With this paper, we give a first reference implementation that hopefully establishing a new direction for improvement for the next few years

110

Lukas Heinrich @lukasheinrich.com · 08/05/2026

can we train one that works for many detectors our of the box? The idea is: try to learn a composable building block (hence LEGO/BRICKS :) ) of next-particle prediction that distills the collective effects of particle interacting with matter that can be iterated just like mechanistic simulators

120

Lukas Heinrich @lukasheinrich.com · 08/05/2026

I'm very excited about this new paper, which kicks of our LEGO ERC project we started earlier this year towards general-purpose AI surrogates for simulating radiation-matter interaction. Instead of training AI surrogates for specific detectors or material distribution ...

1126

Lukas Heinrich @lukasheinrich.com · 12/04/2026

everything new is old

040

Lukas Heinrich @lukasheinrich.com · 01/02/2026

Scaling Laws in Particle Physics Data! This is a result I've been itching to share and it's finally out. One of the big open questions is how much better AI-based methods at particle colliders can still become. 1/4

3167

Lukas Heinrich @lukasheinrich.com · 28/01/2026

An @inspirehep.net HEP mystery: what happened here?

030

Lukas Heinrich @lukasheinrich.com · 05/09/2025

@annalenakofler.bsky.social giving a great talk on SBI in gravitational waves at our „Build Big vs Build Smart Workshop“ indico.ph.tum.de/event/7906/

0171

Lukas Heinrich @lukasheinrich.com · 03/09/2025

1/New paper led by Matthias Vigl arxiv.org/abs/2509.01397 - we wanted to see whether you can see double descent and a benefit of overparametrization in particle physics data and tasks. We see both model- and epoch-wise double descent but the story is more complicated than we thought:

181

Lukas Heinrich @lukasheinrich.com · 09/07/2025

Always love being back at @cern.bsky.social

071

Lukas Heinrich @lukasheinrich.com · 15/12/2024

And next proud advisor moment: @annalenakofler.bsky.social talking about her masters thesis project we worked on while she was still at TUM.

1313

Lukas Heinrich @lukasheinrich.com · 15/12/2024

Proud of Nicole Hartman running the show at the ML for Physical Sciences Workshop at NeurIPS happening today

0274

Lukas Heinrich @lukasheinrich.com · 11/12/2024

so these were actual physical transparencies? What was used to author this? TeX?

140

Lukas Heinrich @lukasheinrich.com · 09/12/2024

It’s a pleasure to host @aishikghosh.bsky.social in Munich, where he he’s sharing the NSBI story that he has pushed in @atlasexperiment.bsky.social for many years and recently came out

Aishik Ghosh giving a seminar at the Max Planck Institute for Physics

292