Machine Learning meets

Many-Body Physics

Many-Body Physics

University of Nottingham, UK, April 8-10, 2024

This workshop will bring together a small number of experts from across the physics/applied-maths/CS divide with common interests in machine learning theory and its applications in order to exchange ideas and identify synergies.

Giulio Biroli (ENS Paris), Marin Bukov (MPI-PKS), Gino Cassella (Imperial), Arnaud Doucet (DeepMind/Oxford), Marylou Gabrié (Polytechnique Paris), Martin Gärttner (Jena), Thiago Guedes (Juelich), Markus Heyl (Augsburg), Zala Lenarčič (Ljubljana), Sirui Lu (MPQ), Florian Marquardt (MPI Erlangen), Nicola Pancotti (AWS), Frank Pollmann (TU Munich), Dominic Rose (UCL), Grant Rotskoff (Stanford), Adam Smith (Nottingham), Riccardo Zecchina (Bocconi)

Accommodation and lectures will be in the Jubilee Campus Hotel at the University of Nottingham. Lectures in meeting room 5A/5B.

- 09:30-10:10 Florian Marquardt (MPI Erlangen): Machine Learning for Quantum Circuits
- 10:10-10:50 Martin Gärttner (Jena): Machine-learning assisted readout of quantum simulators
- 10:50-11:30 Coffee break
- 11:30-12:30 Arnaud Doucet (DeepMind/Oxford): From denoising diffusions to Schrodinger bridges
- 12:30-14:00 Lunch
- 14:00-14:40 Marylou Gabrié (Polytechnique Paris): Assisting sampling of metastable systems with generative models
- 14:40-15:20 Giulio Biroli (ENS Paris): Generative AI and diffusion models: a statistical physics perspective
- 15:20-15:50 Coffee break
- 15:50-16:30 Marin Bukov (MPI-PKS): Self-Correcting Quantum Many-Body Control using Reinforcement Learning withTensor Networks
- 16:30-17:10 Dominic Rose (UCL): Combining reinforcement learning and tensor networks for rare trajectory sampling

- 09:30-10:10 Frank Pollmann (TU Munich): Learning Phases of Matter with Quantum Convolutional Neural Networks
- 10:10-10:50 Adam Smith (Nottingham): Using neural networks to navigate the quantum optimisation landscape
- 10:50-11:20 Coffee break
- 11:20-12:00 Zala Lenarčič (Ljubljana): Reconstructing effective Hamiltonians with unsupervised learning
- 12:00-12:40 Markus Heyl (Augsburg): Solving 2D quantum matter with neural quantum states
- 12:40-14:00 Lunch
- 14:00-14:40 Georg Martius/Igor Lesanovsky (Tübingen): Equation learning and learning of generators
- 14:40-15:20 Riccardo Zecchina (Bocconi): TBA (about Results at the interface between ML and cNeuro)
- 15:20-15:50 Coffee break
- 15:50-16:30 Grant Rotskoff (Stanford): Building data-efficient generative models of proteins with modular sampling strategies
- 16:30-17:10 Nicola Pancotti (AWS): Message passing techniques for tensor networks contractions

- 09:30-10:10 Gino Cassella (Imperial): Freedom from basis sets with Neural Network Variational Monte Carlo
- 10:10-10:50 Sirui Lu (MPQ): Variational Neural and Tensor Network Approximations of Thermal States
- 10:50-11:30 Coffee break
- 11:30-12:10 Thiago Guedes (Juelich): Quantum cellular automata for quantum error correction
- 12:10-14:00 Lunch - End

There are no set dinner plans. The Jubilee Hotel offers a basic dinner service from the central bar. There are not many places near the hotel (just the Rose & Crown, a gastropub, or some takeaways). There are many restaurants in the city centre. Here are some recommendations: (i) In the Lace Market area, Bar Iberico* (tapas), Pizzamisu* (pizza), Sexy Mamma (pasta); (ii) near the Market Square, Kushi Ya* (Japanese), Cosy Club* (British), French Living (bistro), Bill's (fancy burgers); (iii) near Nottingham Castle, Hand & Heart* (gastropub in the Nottingham caves), Fothergills (gastropub), and the cluster of upmarket Indian restaurants on Maid Marian Way, like MemSaab and The Cumin*.

- Juan P. Garrahan (Physics and Astronomy, Nottingham)
- Igor Lesanovsky (Theoretical Physics, Tübingen)
- Georg Martius (Computer Science, Tübingen)
- Hugo Touchette (Mathematical Sciences, Stellenbosch)

This workshop is funded from a network collaboration between the Universities of Nottingham (School of Physics and Astronomy), Tübingen (Institute for Theoretical Physics and Department of Computer Science), and Stellenbosch (Department of Mathematical Sciences).