Reproduce “Learning a Mixture of GFlowNets” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tiago da Silva, Amauri H. Souza, Salem Lahlou
Stratum-Conditioned GFlowNets divide a discrete state space into subsets and train components to sample high-reward objects through a shared index-conditioned model or independent models. In the Lines task with N=256, the method reached comparable fit in fewer than one-third as many training steps as a conventional GFlowNet. Its reachability checks may be intractable for general covers.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.