Id: arxiv/2607.09088 Url: https://arxiv.org/abs/2607.09088 Tags: math-ph, physics.plasm-ph Score: 0 Title: TokaGrad: End-to-end differentiable tokamak simulator for L-to-H full scenario optimization Domain: math-ph Source: arxiv Authors: Jaemin Seo Reviews: 0 Subject: math-ph Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-07-10 Page: https://delphi.science/r/arxiv/2607.09088 As fusion energy moves from theoretical feasibility toward commercialization, design of new reactor concepts, autonomous tokamak control, and high-performance scenario optimization are becoming increasingly important. Traditionally, such optimization tasks have relied on costly trial-and-error or brute-force parameter searches, based on black-box experiments or simulations. Recently, advances in differentiable programming are changing the paradigm of numerical simulation. Unlike conventional simulations, which are typically executed as locally connected step-by-step procedures,…