Harsh Der: 4D Phase Space Reconstruction based on Neural Networks and Differentiable Simulations

Contribution List

3. 4D Phase Space Reconstruction based on Neural Networks and Differentiable Simulations

Harsh Der

29/07/2025, 10:45

Focus report

A neural network-based framework enables precise reconstruction of the four-dimensional phase space of electron beams in particle accelerators. Utilizing a Transformer based Neural Network architecture and differentiable simulations, this approach captures complex features in beam distributions directly from experimental screen images which cannot be captured by measurement techniques used at...

Back to the event