The DEEP-IN Working Group commenced its kick-off meeting on April 23, 2024, with a hybrid gathering of more than 40 participants including 19 members.

The session began with opening remarks from Lingxiao Wang, self-introductions from each member, and a concise introduction to the applications of deep learning for solving inverse problems in sciences. Notable speakers included Akinori Tanaka from RIKEN-AIP/iTHEMS, who gave a vivid introduction to machine learning and his current work, and Gert Aarts from Swansea University, who explored lattice field theories with deep learning, which could also benefit deep learning. Márcio Ferreira introduced the conditional variational auto-encoder(cVAE) for building dense matter equation of states from neutron star observations. Andreas Ipp gave a brief introduction to his work on exploring the early stages of heavy ion collisions and training L-CNNs for lattice gauge theories. The last speaker was Akira Harada, who presented his current work on the application of machine learning to the simulation of supernovae.

During the discussion, members actively brainstormed potential projects and discussed methodologies, emphasizing the importance of interdisciplinary collaboration. The meeting concluded with Tetsuo Hatsuda's optimistic closing remarks about the group's potential to innovate at the intersection of deep learning and physics. There will be more activities from the DEEP-IN Working Group in the near future.

Reported by Lingxiao Wang