DIFFERENTIABLE PROGRAMMING FOR PARTICLE PHYSICS SIMULATIONS

نویسندگان

چکیده

We describe how to apply adjoint sensitivity methods backward Monte-Carlo schemes arising from simulations of particles passing through matter. Relying on this, we demonstrate derivative based techniques for solving inverse problems such systems without approximations underlying transport dynamics. are implementing those algorithms various scenarios within a general purpose differentiable programming C++17 library NOA (github.com/grinisrit/noa).

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Differentiable Genetic Programming

We introduce the use of high order automatic differentiation, implemented via the algebra of truncated Taylor polynomials, in genetic programming. Using the Cartesian Genetic Programming encoding we obtain a high-order Taylor representation of the program output that is then used to back-propagate errors during learning. The resulting machine learning framework is called differentiable Cartesia...

متن کامل

Anniversaries for particle physics.

his month and next month mark two important anniversaries in the field of particle physics— the 50th anniversary of the inauguration of Fermilab (the Fermi National Accelerator Laboratory, just outside of Chicago, Illinois) and the 5th anniversary of the Higgs boson discovery at CERN (the European Organization for Nuclear Research, in Geneva, Switzerland). These two institutions have been unite...

متن کامل

Regularity Conditions for Non-Differentiable Infinite Programming Problems using Michel-Penot Subdifferential

In this paper we study optimization problems with infinite many inequality constraints on a Banach space where the objective function and the binding constraints are locally Lipschitz‎. ‎Necessary optimality conditions and regularity conditions are given‎. ‎Our approach are based on the Michel-Penot subdifferential.

متن کامل

A Differentiable Physics Engine for Deep Learning in Robotics

One of the most important fields in robotics is the optimization of controllers. Currently, robots are often treated as a black box in this optimization process, which is the reason why derivative-free optimization methods such as evolutionary algorithms or reinforcement learning are omnipresent. When gradient-based methods are used, models are kept small or rely on finite difference approximat...

متن کامل

Java simulations for physics education

We discuss the use of World Wide Web-based Java simulations in teaching physics to K{12 and undergraduate students. Our work focuses on the physics of membranes and illustrating how such systems are studied. We propose that Java should be used not only to produce small versions of research simulations but also to provide models illustrating simpler concepts underlying the main subject matter. I...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: ????. ?????? ? ????????

سال: 2022

ISSN: ['0044-4510']

DOI: https://doi.org/10.31857/s0044451022020043