Employer: Triumf
Work term duration: 4 months
Location: Vancouver, BC (UBC campus)
Job Description
Discover TRIUMF, Canada’s particle accelerator centre, and one of the world’s leading laboratories for particle and nuclear physics and accelerator-based science. TRIUMF’s diverse community of nearly 600 multidisciplinary researchers, engineers, technicians, tradespeople, staff, and students create a unique incubator for Canadian excellence, as well as a portal to premier global collaborations. Fueling innovation and improving lives, we are committed to accelerating discovery and shaping a better world.
Equity, diversity, and inclusion are integral to excellence and enhance our ability to create knowledge and opportunity for all. Together, we are committed to building an inclusive culture that encourages, supports, and celebrates the voices of our employees, students, partners, and the people and communities we serve.
TRIUMF’s student program typically hires ~40 students each term from across all disciplines who hang together socially and professionally during the term, and beyond. Our program offers young minds a chance to gather, learn, make new friends, and have fun doing so during their time at TRIUMF. We hold weekly Pro-D sessions for students which include different activities such as conflict resolution, emotional intelligence, critical thinking, and active listening. Additionally, we hold weekly seminars about the different experiments at TRIUMF, and these talks specifically cater to students. Located on the UBC campus makes it easier for students to make friends, create memories, and explore Vancouver all while working at TRIUMF.
TRIUMF’s student program is more than just a place to work, its learning, memories, friendships, exploration, and development all clubbed into one.
Come join us to see how we are working to unlock the mysteries of the universe and shape the future. Apply now and be part of our extraordinary journey.
Learn more about the amazing research and work we do at TRIUMF.
Overview
Particle physics experiments at the Large Hadron Collider (LHC) study the fundamental building blocks of the Universe and interactions between them. TRIUMF scientists are leading efforts in multiple aspects of the ATLAS experiment at the Large Hadron Collider including detector development, data acquisition and data analysis including machine learning applications. During the High Luminosity run of the LHC (anticipated to start in the late 2020s) ATLAS experiment will enter a new era of Higgs Boson precision studies, and searches for physics Beyond the Standard Model. This effort will require several of millions of CPU-years per year, mostly spent in providing simulation and reconstruction of simulated data. Deep Learning methods are already becoming a critical part of such simulations and LHC data analysis.
TRIUMF’s team of scientists, together with their national and international partners, developed a method combining ‘classical’ Deep Learning methods with the unique capabilities of quantum annealing processors to create a quantum variational autoencoders (QVAEs) generating simulated outputs such as those of the calorimeters of the ATLAS experiment.
During the course of the project the student will utilize ATLAS calorimetric cluster data and generic simulated electromagnetic calorimeter data to further develop the model, aiming to increase it’s performance.
Successful completion of the project would have outsize impact on the field of particle physics, most notably on the study of the Higgs Boson.
Duties
– Development and adaptation of machine learning and quantum machine learning techniques for generation of synthetic calorimeter samples (calorimetric simulation)
– Training and tuning of the models
– Deployment on a quantum annealing processor
– Tuning of annealing parameters
– Evaluation of the performance of the developed models
– Collaboration with particle physicists and machine learning experts from industry and academia
– Presentations, oral and written reports to peers and experts in the field
Skills Learned During This Work Experience
– Fundamental and technical expertise in generative methods, particularly for quantum annealing processors
– Critical thinking and problem solving
– Expert level knowledge of deep learning libraries
– Oral and written presentation skills; Scientific communication
Qualifications:
The position targets a senior undergraduate in an Joint Physics and Computer Science, Engineering Physics, Computer Science, Computer and Electrical Engineering or similar program.
Ideally the candidate will have completed at least two work terms.
Qualifications required
– Strong knowledge of physics curriculum at senior undergraduate level.
– Strong background in python programming.
– Experience with numerical and machine learning libraries in python: numpy, matplotlib, scikit-learn
– Strong knowledge of statistics and machine learning concepts.
– Strong knowledge of deep learning concepts, particularly generative models (e.g., RBMs, VAEs, Transformers).
– Strong technical background in at least one deep learning library pytorch, tensorflow
Beneficial experience
– Experience with versioning tools (git)
– Experience with batch processing systems (slurm, torque)
– Experience with jupyter
– Experience running container services (docker/apptainer)
Skills required:
Autoencoders, Data Analysis, Deep Learning, Git, Jupyter Notebook, Machine Learning, Physics, Problem Solving, PyTorch, Slurm Workload Manager, Statistics