Local Search GFlowNets
This addresses a specific bottleneck in GFlowNets for generating high-reward samples in domains like biochemistry, representing an incremental advancement.
The paper tackles the issue of GFlowNets sometimes failing to produce high-reward samples due to over-exploration by introducing local search to exploit high-rewarded areas, resulting in significant performance improvements in biochemical tasks.
Generative Flow Networks (GFlowNets) are amortized sampling methods that learn a distribution over discrete objects proportional to their rewards. GFlowNets exhibit a remarkable ability to generate diverse samples, yet occasionally struggle to consistently produce samples with high rewards due to over-exploration on wide sample space. This paper proposes to train GFlowNets with local search, which focuses on exploiting high-rewarded sample space to resolve this issue. Our main idea is to explore the local neighborhood via backtracking and reconstruction guided by backward and forward policies, respectively. This allows biasing the samples toward high-reward solutions, which is not possible for a typical GFlowNet solution generation scheme, which uses the forward policy to generate the solution from scratch. Extensive experiments demonstrate a remarkable performance improvement in several biochemical tasks. Source code is available: \url{https://github.com/dbsxodud-11/ls_gfn}.