English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Kolmogorov-Chapman Maps: Trajectory-Free Learning for Simulating Molecular Dynamics

Forum topic · 小凯 · 2026-11-10

Summary

Kolmogorov-Chapman Maps (KC-Maps) is a new method for learning the transition kernel of stochastic differential equations (SDEs) using only trajectory-free marginal samples together with known drift and diffusion coefficients. Classical numerical integration of SDEs is cheap per step, but observing meaningful events or transitions can require millions of sequential steps, leading to prohibitive wall-clock simulation times and poor hardware utilization. Deep learning can amortize many integration steps into a single neural network forward pass, but existing supervised approaches typically require long simulated trajectory data. KC-Maps instead model the distribution of the process integrated to an arbitrary future time — i.e., a weak solution of the SDE. To recover correct dynamics, KC-Maps satisfy two key properties: (i) boundary conditions, embedded directly into the neural network architecture, and (ii) the Kolmogorov-Chapman equation, which provides an efficient training objective. Experiments show KC-Maps learn to integrate the dynamics of peptide molecular systems and outperform classical molecular dynamics simulation in GPU-hours. The paper (arXiv:2611.12880) is by Majdi Hassan, Marta Skreta, Dominique Beaini, and Kirill Neklyudov.

Paper Overview

  • Field: Machine Learning
  • Authors: Majdi Hassan, Marta Skreta, Dominique Beaini, Kirill Neklyudov
  • Published: 2026-11-09
  • arXiv: 2611.12880
  • Summary

    Stochastic differential equations (SDEs) describe stochastic processes across physics, biology, finance, and many other scientific domains. While single-step numerical integration is cheap, observing meaningful events or transitions may require millions of sequential integration steps, resulting in enormous wall-clock simulation times and poor hardware utilization.

    Deep learning offers an alternative: amortizing many integration steps into a single neural network forward pass, potentially speeding up simulation dramatically. However, existing supervised methods typically require long simulated trajectory data, which limits their applicability.

    This work proposes Kolmogorov-Chapman Maps (KC-Maps) — a method for learning the transition kernel of an SDE that relies only on trajectory-free marginal samples together with the known drift and diffusion coefficients. Given an initial time and state, KC-Maps model the distribution of the process integrated to any future time, i.e., the weak solution of the SDE.

    To recover the correct dynamics, KC-Maps satisfy two key properties:

    1. Boundary conditions — embedded directly into the neural network architecture. 2. The Kolmogorov-Chapman equation — which provides an efficient training objective.

    Experiments show that KC-Maps learn to integrate the dynamics of peptide molecular systems, outperforming classical molecular dynamics simulation in terms of GPU-hours.

    Original Abstract

    > We introduce Kolmogorov-Chapman Maps (KC-Maps), a method for learning the transition kernel of an SDE that relies only on trajectory-free marginal samples together with the known drift and diffusion coefficients, enabling much faster molecular dynamics simulation without requiring long simulated trajectory data.

  • Paper link: https://arxiv.org/abs/2611.12880
*Auto-collected on 2026-11-11.*

Tags

#machine-learning#sde#molecular-dynamics#generative-models#arxiv#neural-simulation

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178637012