AI RESEARCH
ReMAP: Neural Reparameterization for Scalable MAP Inference in Arbitrary-Order Markov Random Fields
arXiv CS.LG
•
ArXi:2411.18954v4 Announce Type: replace Scalable high-quality MAP inference in arbitrary-order Marko Random Fields (MRFs) remains challenging. Approximate message-passing methods are often efficient but can degrade on dense or high-order instances, while exact solvers such as Toulbar2 become increasingly expensive at scale. We present ReMAP, an instance-wise neural reparameterization framework that directly optimizes a differentiable relaxation of the original MRF energy. Instead of relying on supervised labels or amortized.