Monte Carlo Seminar| Guan-Horng Liu|Sampling with Schrödinger Bridge—An Adjoint-Matching Perspective

Online Monte Carlo Seminar
sites.google.com/view/monte-carlo-seminar

Speaker: Guan-Horng Liu (Meta AI)

Title: Sampling with Schrödinger Bridge—An Adjoint-Matching Perspective

Abstract: Computational methods for learning to sample from the Boltzmann distribution—where the target distribution is known only up to an unnormalized energy function—have advanced significantly recently. Due to the lack of explicit target samples, prior diffusion samplers often require importance-weighted estimation or complicated learning processes. Both trade off scalability with extensive evaluations of the energy and model, thereby limiting their practical usage. In this talk, I will introduce Adjoint Schrödinger Bridge Sampler (ASBS), a new diffusion sampler that employs simple and scalable matching-based objectives yet without the need to estimate target samples during training. ASBS is grounded on a mathematical model—the Schrödinger Bridge—which enhances sampling efficiency via kinetic-optimal transportation. Through a new lens of stochastic optimal control theory, we demonstrate how SB-based diffusion samplers can be learned at scale via Adjoint Matching and prove convergence to the global solution. We demonstrate the effectiveness of ASBS on sampling from classical energy functions, amortized conformer generation, and molecular Boltzmann distributions.