The Score Hamiltonian: Mapping Diffusion Models to Adiabatic Transport

Peter Halmos; Boris Hanin

math-phMay 28, 2026

The Score Hamiltonian: Mapping Diffusion Models to Adiabatic Transport

Authors: Peter Halmos, Boris Hanin

Organizations: Computer Science Department, Princeton University. · ORFE Department, Princeton University.

Abstract

We exhibit an exact correspondence between sampling with score-based diffusion models and adiabatic transport of ground states for a family of Schrödinger operators we call Score Hamiltonians, built from the learned score's quantum potential. We obtain novel density reconstruction bounds and principled annealing schedules via adiabatic theorems for Fokker-Planck equations with time-varying potentials. We find the fundamental limit of sampling is set by the ratio of squared score-matching error to Score Hamiltonian spectral gap - the inverse Poincaré constant of the data density.

Explore similar work

Diffusion Models Observe Only Gradients: A Geometric Perspective on Score Matching Errors
Jun 4, 2026Naïl B. Khelifa, Richard E. Turner, Ramji VenkataramananScore-Based Diffusion Model Kullback-Leibler Divergence
From Score Learning to Discretized Sampling: An End-to-End Generalization Analysis of Diffusion Models
Jul 25, 2026Jinshu Huang, Yiming Jiang, Chunlin WuScore-Based Diffusion Model Generative
Low-dimensional adaptation of diffusion models: Convergence in total variation
Jan 22, 2025Jiadong Liang, Zhihan Huang, Yuxin ChenScore-Based Diffusion Model Denoising Diffusion Probabilistic Model

Jun 4, 2026 · stat.MLJ/K move · Enter open · S save

Diffusion Models Observe Only Gradients: A Geometric Perspective on Score Matching Errors

Naïl B. Khelifa, Richard E. Turner, Ramji Venkataramanan

University of Cambridge

Score-based diffusion models are typically trained by minimizing the

L^2

score matching error, and standard theoretical analyses rely on this quantity to bound the sampling discrepancy between the learned and target distributions. We show the

L^2

score error is not the right intrinsic measure of marginal distributional quality: a learned diffusion model can incur arbitrarily large

L^2

score error while perfectly matching the target distribution. By decomposing score errors into a gradient and a solenoidal component (a Helmholtz-Hodge decomposition), we identify the geometric reason behind this: only the gradient component enters the marginal Fokker-Planck dynamics, while the solenoidal component is structurally invisible. We make this precise in three results. First, building on the corrected geometry, we prove an impossibility result: no monotone function of the

L^2

score error can uniformly lower bound any divergence between the learned and target distributions. Second, we derive an upper bound on the Kullback-Leibler divergence that depends only on the observable gradient component of the error, tightening the standard Girsanov bound for generic score networks, and identifying its looseness as the cost of operating on path-space rather than marginal-space dynamics. Third, we give a tractable estimator of the gradient component via a dual Sobolev identity, which is shown to empirically correlate substantially better with sample quality than the full

L^2

error.

Score-Based Diffusion Model Kullback-Leibler Divergence Gradient Concentration Fokker--Planck Equation

Open paper Ask about it PDF

Abstract

Explore similar work

Diffusion Models Observe Only Gradients: A Geometric Perspective on Score Matching Errors

From Score Learning to Discretized Sampling: An End-to-End Generalization Analysis of Diffusion Models

Low-dimensional adaptation of diffusion models: Convergence in total variation