Senior Researcher @know_center, working in machine learning.
Graz, Österreich
Joined December 2017
- Tweets322
- Following184
- Followers178
- Likes1.1K
Pinned Tweet
If you want to visit me at @Know_Center to work in the fields of information theory for deep learning or physics- informed neural networks, this might be a great opportunity. Drop me a line if you are interested!
Exciting News!
🌟The Open Call for ENFIELD's Third-party Exchange Scheme (TES) is now live 🚀
🗓️ Feb 1 - Mar 31, 2024 💡 Addressing AI challenges in Healthcare, Energy, Manufacturing, and Space
🔗 Details & Apply: enfield-project.eu/1st_open_…
ENFIELD Partner Institutions cannot apply!
Bernhard Geiger retweeted
On March 11-12, we gathered in #Brdo_pri_Kranju, #Slovenia, for a workshop on #AI in sustainable energy & #GreenAI, hosted by #Iskraemeco & @Know_Center
🔍AI-driven energy optimization
🔧Predictive maintenance
🔋Smart energy storage
🌐AI in energy forecasting
💡Sustainable AI
Feel free to reach out if you want to join @FRohrhofer and me at @Know_Center for work on physics-informed machine learning!
Bernhard Geiger retweeted
🌟3rd Exchange Scheme Open Call🌟
We invite PhD students, post-docs, and senior researchers to propose innovative projects addressing #ENFIELD challenges and advancements in #AI research areas:
*Adaptive AI
*Green AI
*Human-Centric AI
*Trustworthy AI
Join us in making an impact
Bernhard Geiger retweeted
🚀Exciting times ahead for #ENFIELD🌍
We’re thrilled to welcome all our partners to Budapest for the General Assembly III 🎉
The preparation is in full swing – from setting up the room 🖥️ to ensuring everything is ready for productive discussions.
Can’t wait to see everyone here!
Bernhard Geiger retweeted
Join us at the Workshop on AI for Sustainable Energy Systems and Green AI hosted by the HEU project ENFIELD!
📅 11-12 March 2025
Explore how AI can revolutionize energy systems while reducing environmental costs 🚀
📑 Submit your abstract by 10 Feb 2025
enfield-project.eu/ws_ai_sus…
Bernhard Geiger retweeted
Advancing #AIResearch at KC: @LinaraAdylova (@ruhrunibochum) spent 3 months with us through an @HorizonEnfield open call to explore information-theoretic properties of #NeuralNetworks, #LatentSpaceGeometry & generalization performance to enhance our understanding of #DeepLearning
Bernhard Geiger retweeted
❗️ Horizon Enfield's 2nd Exchange scheme Open Call deadline extended to 31th Ocotber 2024. 23:00 CET. More info: enfield-project.eu/oc2_TES_2…
Bernhard Geiger retweeted
The #ENFIELD General Assembly successfully wrapped up in Eindhoven on 11-10 September! 🎉 A big thank you to all the partners for presenting their progress and future plans, and special thanks to TU/e for organizing such an insightful and productive event.
🤯 Our work (arxiv.org/abs/2402.09090) on epsilon-machines, lumpability, and emergence has been covered on Youtube -- thanks so much @skdh for this amazing summary!
1/3 "Lumpability is now a technical term". Take-home from @skdh's wonderful video recap of our preprint on identifying and quantifying emergence (FYI .@seanmcarroll). Thanks Sabine 🙏🏽 - and great stuff project leads @_fernando_rosas & @PedroMediano youtube.com/watch?v=Tw9sr05V…
Very nice article about our recent preprint on #arxiv: arxiv.org/abs/2402.09090; thanks, @QuantaMagazine !
How emergence happens has long been debated amongst physicists, biologists and philosophers alike. A new framework identifies the hierarchical structure that’s key to how order emerges on macroscopic scales. @philipcball reports: quantamagazine.org/the-new-m…
Bernhard Geiger retweeted
𝗘𝗺𝗲𝗿𝗴𝗲𝗻𝗰𝗲 𝗱𝗲𝗺𝘆𝘀𝘁𝗶𝗳𝗶𝗲𝗱!
Great piece by @philipcball on the work by @_fernando_rosas et al. with new methods to formalize emergence. Huge need to get math to back up hypotheses of emergence.
quantamagazine.org/the-new-m…
Bernhard Geiger retweeted
🚀 Join us for the KNOWSKITE-X 2nd Webinar! Discover topics like steel design, ML applications, reinforcement learning for materials design, and much more! Register for free!
knowskite-x.eu/knowskite-x-2… #MachineLearning #MaterialsDiscovery
Bernhard Geiger retweeted
🎉Reflecting on a successful WP2-WP3 Workshop held on April 23-24 at @IMTFrance
🔍We explored pillars and verticals’ use cases, collaboration between WP2 & WP3, and new planning for future TES & TIS Open Calls.
Thank you to all who participated and contributed! 🌟🌟
#ENFIELD #AI
This project is led by @ecoplus_noe and executed jointly with @fh_stpoelten FOTEC @Fraunhofer @wu_vienna IMG and @Know_Center
Proud to initiate #LogiTRAAK! Joining forces with top research institutes & industry leaders to elevate Austrian industry with Asset Tracking & Tracing Systems. Leading #dataanalysis & modeling, along with providing Data Service Cards. Together, we're shaping future #logistics!
Reach out to me if you want to join me and @Know_Center for work on physics-informed ML or information-theoretic analyses of deep learning!
🚨 Deadline Alert!
The 1st Open Call for Horizon Enfield has been extended to April 14th, 2024, 23:00 CET.
Don't miss out! 🌟 #HorizonEnfield #ai #researchers
Don't forget to apply if you are interested in working with me on PINNs or information-theoretic analyses of deep learning!
We are part of the @HorizonEnfield project, which offers €1.1 million to 76 motivated researchers, promoting interdisciplinary AI research and excellence. Join and address topics in Adaptive, Green, Human-Centric, and Trustworthy AI.
🎯enfield-project.eu/1st_open_…
Bernhard Geiger retweeted
Preprint time:
“Software in the natural world: A computational approach to emergence in complex multi-level systems”
arxiv.org/abs/2402.09090
A rigorous framework that not only identifies *when* emergence takes place, but explains *how* this happens computationally
Preprint 📜:
If the world is (microscopically) a Markov chain, and if coarse grainings of Markov chains are usually not Markov, then why do so many (macroscopic) phenomena appear Markov to us?
🧵
(We also discuss how these kinds of closure are connected to the lumpability of Markov chains, the rare scenario where a function of a Markov chain is Markov.)
Joint work with @_fernando_rosas, @loopyluppi, @anilkseth, Daniel Polani, Michael Gastpar, @PedroMediano