AI and Quantum Information for Particle Physics

14 Nov 2023, 09:00 → 17 Nov 2023, 21:50 Asia/Seoul

Minho SON (Korea Advanced Institute of Science and Technology), Myeonghun Park (Seoul National University of Science and Technology)

Particle physics data is unique and challenging due to its tremendous size and complicated structure stemming from its quantum nature. Rapidly improving quantum-computing technologies could be applied within the particle physics in order to help tackling such computing challenges. The goal of the workshop is to discuss the latest developments in the field of machine learning, quantum computation, and their application in high energy physics, and to brainstorm possible collaborative projects. We hope to bring together experts to better understand this intersection and reach the full potential of quantum computing and machine learning.

poster.pdf

Tuesday, 14 November

10:00 → 12:30 Quantum Information

Conveners: Myeonghun Park (Seoul National University of Science and Technology), Joonwoo Bae (KAIST)

10:00 Welcome

Speaker: Minho SON (KAIST)

10:10 Entanglement Theory

Entanglement, that is, quantum correlations that do not have a classical counterpart, is a resource for quantum information processing. I provide an overview on the entanglement theory by focusing on the structure of entangled states. I also discuss the experimental verification of entangled states given assumptions made on sources and measurements.

Speaker: Dr Joonwoo Bae (KAIST)

11:10 Coffee & Discussion

11:30 Quantum Position Verification and Time-Constrained Nonlocal Computation

Quantum position verification (QPV) is a cryptographic task in which the spatial location of an untrusted agent is certified using the principles of quantum mechanics and special relativity. The problem of QPV has deep connections to computational complexity and the AdS/CFT correspondence. In this talk I will introduce the general task of QPV and review some results. I will then turn to recent theoretical work analyzing the structure of QPV protocols in which the distribution of product states is used to certify a spatial location, and an honest prover must perform a joint measurement on the signals. This particular class of QPV protocols reveals separations in security based on whether the adversaries are restricted to classical versus quantum communication.

Speaker: Dr Eric Chitambar (UIUC)

KAIST-2023-QPV-public.pdf

12:30 → 14:00 Lunch

14:00 → 17:30 Quantum Computing

Convener: Dr Jesse Stryker (Lawrence Berkeley National Laboratory)

14:00 Theory-driven Quantum Machine Learning for HEP

Machine Learning is, in most cases, powerful but a black-box application. In this talk, we will tackle this very problem from a quantum mechanics point of view, arguing that an optimisation problem, such as classification or anomaly detection, can be studied by “rephrasing" the problem as a quantum many-body system or a mixed state. Such an approach allows us to employ the entire arsenal of quantum theory for data analysis techniques. Hence, this talk will present a small step towards fully theory-driven and interpretable quantum machine learning applications.

Speaker: Dr Jack Y. Araz (Jefferson Lab.)

jackaraz_kaist.pdf

15:00 Coffee & Discussion

15:30 Digitizaiton and Propagation in quantum computing for lattice gauge theories

Speaker: Dr Ying-Ying Li (University of Science and Technology of China)

Wednesday, 15 November

10:00 → 12:30 Quantum Computing

Convener: Prof. Yingying Li (USTC)

10:00 Combinatorial optimization using QC for collider physics

Speaker: Myeonghun Park (Seoul National University of Science and Technology)

11:00 Coffee & Discussion

11:30 Expressing non-Abelian gauge-field dynamics in the quantum age

Speaker: Dr Jesse R. Stryker (LBNL, Berkeley)

stryker_kaist_2023-11-15.pdf

12:30 → 14:00 Lunch

14:00 → 16:30 Machine Learning

Convener: Lukas Heinrich (echnical University Munich)

14:00 Constructing Novel Nonparametric Estimators for f-divergences and Its Applications to High-energy Physics

Nonparametric methods, such as nearest neighbor and kernel methods, can offer simple and parallelizable algorithms without the need for manual structure tunning. However, these methods may suffer from severe performance degradation due to biases from the high-dimensionality of data. I will introduce recently derived equations for understanding and addressing the high-dimensional bias and present some principled algorithms. These algorithms can be used to estimate the f-divergences, which can be employed in objective functions to select important features or decorrelate information that should be irrelevant. Examples of the applications in high-energy physics and other fields will be provided.

Speaker: Dr Yung-Kyun Noh (Hanyang University)

2023_AI_Quantum_particle_physics_Daejeon_distr.pdf

15:00 Coffee & Discussion

15:30 Scalar and Grassmann Neural Network Field Theories

Neural Networks (NN), the backbones of Deep Learning, define field theories through output ensembles at initialization. Certain limits of NN architecture give rise to free field theories via Central Limit Theorem (CLT), whereas other regimes give rise to weakly coupled, and non-perturbative field theories, via small, and large deviations from CLT, respectively. I will present a systematic construction of free, weakly interacting, and non-perturbative field theories by tuning different attributes of NN architectures, bringing in methods from statistical physics, and a new set of Feynman rules. Some interacting field theories of our choice can be exactly engineered at initialization, by parametrically deforming distributions over stochastic variables in NN architectures. As an example, I will present the construction of $λφ^4$ scalar field theory via statistical independence breaking of NN parameters in the infinite width limit. Lastly, I will introduce free and interacting regimes in Grassmann field theories defined via initialized Grassmann NN architectures.

Speaker: Dr Anindita Maiti (Perimeter Institute)

18:30 → 21:00 Workshop dinner

Thursday, 16 November

10:00 → 12:30 Machine Learning

Convener: Anindita Maiti (Perimeter Institute)

10:00 Towards Building Large HEP Models with Self-Supervised Learning

Speaker: Dr Michael A. Kagan (SLAC)

11:00 Coffee & Discussion

11:30 The Vision of End-to-End ML models in HEP

Speaker: Dr Lukas Heinrich (echnical University Munich)

12:30 → 14:00 Lunch

14:00 → 15:00 Machine Learning

Convener: Michael A. Kagan (SLAC)

14:00 Why autoencoders fail at anomaly detection and what we can do about it

Speaker: Dr Sangwoong Yoon (KIAS-AI)

20231116_swyoon_anomaly_detection_share.pdf

15:00 → 17:30 Quantum Computing

Convener: Jack Y. Araz (Jefferson Lab.)

15:00 Quantum computing for high-energy physics

Speaker: Dr Michael Spannowsky (University of Durham)

16:00 Coffee & Discussion

Friday, 17 November

10:00 → 11:50 Machine Learning

Convener: Minho SON (Korea Advanced Institute of Science and Technology)

10:00 Mixture Density Network for Neutrino Reconstruction in Collider Physics

Speaker: Dr SeungJin Yang (Kyung Hee University)

10:40 Exploring local and global feature integration in Multi-Model Deep Neural Networks

Speaker: Ms Kayoung Ban (Yonsei University)

11:20 Coffee & Discussion

11:50 → 12:30 Quantum Computing

Convener: Myeonghun Park (Seoul National University of Science and Technology)

11:50 Quantum Amplitude Amplification Operators: Quantum Search in the NISQ era

Speaker: Hyeokjea Kwon (KAIST)

12:30 → 14:00 Lunch

14:00 → 16:30 Future Prospective

Conveners: Minho SON (Korea Advanced Institute of Science and Technology), Myeonghun Park (Seoul National University of Science and Technology)

14:00 TBA

Speaker: Minho SON (Korea Advanced Institute of Science and Technology)

15:00 Coffee and Discussion

15:30 Summary and Future prospective

Speakers: Minho SON (Korea Advanced Institute of Science and Technology), Myeonghun Park (Seoul National University of Science and Technology)

16:30 → 17:00 Break

17:00 → 18:00 Discussion and collaboration

Conveners: Minho SON (Korea Advanced Institute of Science and Technology), Myeonghun Park (Seoul National University of Science and Technology)