AI has 'learned' the fundamental rules of how plasma particles collide, discovering patterns that have eluded human physicists for nearly a century.
March 31, 2026
Original Paper
From molecular dynamics to kinetic models: data-driven generalized collision operators in 1D3V plasmas
arXiv · 2603.27828
The Takeaway
Scientists have long relied on simplified, 90-year-old formulas to predict plasma behavior, but they often fail in the extreme conditions of fusion reactors. By training a model on the raw movements of millions of particles, researchers found a new set of 'generalized' collision rules that could be the key to finally achieving stable nuclear fusion.
From the abstract
We present a data-driven approach for constructing generalized collisional kinetic models for inhomogeneous plasmas in one-dimensional physical space and three-dimensional velocity space (1D-3V). The collision operator is directly learned from micro-scale molecular dynamics (MD) and accurately accounts for the unresolved particle interactions over a broad range of plasma conditions. Unlike the standard Landau operator, the present operator takes an anisotropic, non-stationary form that captures