A one-loop correction, computed from two auxiliary ODEs, brings deterministic generative samplers close to the stochastic reference (53% error reduced to 1.6% on a cubic drift), within a path-integral framework that unifies flow, diffusion, variational, and adversarial models.
Weinberg,Phenomenological Lagrangians, Physica A96(1979) 1-2,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Unifying Generative Models with Path Integrals
A one-loop correction, computed from two auxiliary ODEs, brings deterministic generative samplers close to the stochastic reference (53% error reduced to 1.6% on a cubic drift), within a path-integral framework that unifies flow, diffusion, variational, and adversarial models.