(Conditional) generative adversarial networks (GANs) have had great success in recent years, due to their ability to approximate (conditional) distributions over extremely high-dimensional spaces.

In this paper we evaluate the conditional Yeh-Wiener integral E(F(x)∣ x(s, t) = ξ) for functions F of the form F(x) = exp{∫t 0 ∫s 0 φ(σ, τ, x(σ, τ)) dσ ...