Lattice Random Walk Discretisations of Stochastic Differential Equations

Samuel Duffield, Maxwell Aifer, Denis Melanson, Zach Belateche, Patrick J. Coles · 2025

Read Lattice Random Walk Discretisations of Stochastic Differential Equations by Samuel Duffield, Maxwell Aifer, Denis Melanson, Zach Belateche, Patrick J. Coles (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-20883.txt.

Description

Samuel Duffield and four co-authors replace smooth random steps with coin flips in equations of chance, in a 2025 paper. Stochastic differential equations model everything from stock jitter to AI image generation. Simulating them usually needs floating-point arithmetic and Gaussian random numbers at every step.

Their scheme walks on a lattice, sampling one or two random bits per step to pick each move. The crude steps still converge to the right distributions, which they prove. The approach shrugs off rounding errors and even handles drifts that break standard assumptions. It also fits hardware that manipulates probability streams directly instead of numbers. Tests run from textbook equations to the diffusion models behind modern image tools.

Catalog record

CREATORSamuel Duffield, Maxwell Aifer, Denis Melanson, Zach Belateche, Patrick J. Coles
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-20883

Browse the full library · About the Tower of Babel