← Search

Xiaochen Du

2 accepted papers

2026

Discrete Adjoint Schrödinger Bridge Sampler

ICML 2026poster

Learning discrete neural samplers is challenging due to the lack of gradients and combinatorial complexity. While stochastic optimal control (SOC) and Schrödinger bridge (SB) provide principled solutions, efficient SOC solvers like adjoint matching (AM), which excel in continuous domains, remain une…

Cited by 0SourceScholar
2026

MetaDNS: Enhancing Exploration in Discrete Neural Samplers via Metadynamics

ICML 2026poster

Sampling from discrete distributions with multiple modes and energy barriers is fundamental to machine learning and computational physics. Recent discrete neural samplers like MDNS suffer from mode collapse and fail to sample high-energy barrier regions between modes, which is critical for free ener…

Cited by 0SourceScholar