@PilaRicoCastro @FECYT_Ciencia @craftoa_project @DelipaltaAlex @RDA_Europe @elli_lib @OSTrails @rich_ncps @EU_Commission @CienciaGob @druiziruela @pit_peschke @ansantam @KITKarlsruhe @AI4EOSC @AnnikaThies @eRImote_EU @habermann_k @canSERV_EU Jose Luis Martínez, coordinador de @ESFRI_eu, interviene antes del cierre institucional de la jornada #ISCIII sobre Infraestructuras Europeas de Investigación. Streaming ➡️ https://t.co/eTLLGZSxsX

#RL4AA24 has been a great experience! I was able to talk about my research at @IFMIF_DONES and learn more about autonomous accelerators. Congratulations to the organisers and long live to #ReinforcementLearning! 🤖❤️

This Q1 our team is really on the road teaching and giving machine learning tutorials all over! 👩🏻‍🏫 It was a pleasure to come back to the Deep Learning School organized by @ErUMDataHub and give an introduction course about neural networks

On February 5-7 2024 we hosted the 2nd Reinforcement Learning for Autonomous Accelerators in Salzburg 🎉 ! It was splendid to see how this community is growing, and I am delighted that we could provide some interesting content, including the tutorial on meta RL.

I think the paper is a great primer to started and has very interesting details on the advantages and challenges of applying it to accelerators. Have a look!

While this method is mostly known in machine learning to tune the parameters of other algorithms, it is extremely useful to directly tune the parameters of accelerators

We recently submitted a review paper on applying Bayesian optimization methods in particle accelerators! https://t.co/Prj1h507Sz

Join us next year to learn about neural networks! I'll be teaching the introductory lecture :)

If you are interested in knowing more about machine learning applied to accelerators you can have a look at our new publication: https://t.co/Mbux16mBtU   This is just a teaser, review papers are coming 📖

We have a new postdoc position open to work with Prof. De Carne and me in the field of machine learning and energy efficiency for accelerators. Have a look! https://t.co/174cRjwARk

@helmholtz_ai @HIDAdigital Kudos also to the team members for the great work (despite not regularly working with CNNs) @crkesu

It was really interesting to be exposed to the actual energy that training large models require, and we came back with a long list of strategies to reduce it 🤓 Big thanks to the organizers, we had a lot of fun (and learned a lot too)! @helmholtz_ai @HIDAdigital

As researchers we strived for the best accuracy (and our model was indeed the most accurate!) but at the price of an increased energy consumption.

The caveat was to maintain accuracy while minimizing the energy needed to develop and train the model (sent to the GPUs of the HoreKa supercomputer facility) and also minimizing the energy for inference.

The AI4Accelerators team of KIT participated in the AI Hero Hackathon 2023!! We joined the research field Energy, and the goal was to tune a Mask R-CNN model to identify thermal leaks from images of buildings in Karlsruhe

Our new paper on Bayesian optimization to improve the injection efficiency of a particle accelerator is out! Check it out https://t.co/QtIq9lbt7q @crkesu

Special thanks to @crkesu and @jank324 for their hard work!!

We hosted it at @KITKarlsruhe at the Institute of Beam Physics and Technology, where a very nice community of reinforcement learning enthusiasts from all around the world met to showcase their work and discuss the present and future of the field

I am so grateful for all the people that came together at the 1st collaboration workshop on Reinforcement Learning for Autonomous Accelerators!

Some pics from the AI Hub Karlsruhe event! I loved discovering the huge potential of ML in other fields (such as healthcare, brilliant keynote by @AnalogAldo, food for thought!). It was also fun to discover that everyone loves Bayesian optimization, regardless of the field 🤓