Maximum entropy GFlowNets with soft Q-learning

arXiv:2312.14331v2 Announce Type: replace
Abstract: Generative Flow Networks (GFNs) have emerged as a powerful tool for sampling discrete objects from unnormalized distributions, offering a scalable alternative to Markov Chain Monte Carlo (MCMC) methods. While GFNs draw inspiration from maximum entropy reinforcement learning (RL), the connection between the two has largely been unclear and seemingly applicable only in specific cases. This paper addresses the connection by constructing an appropriate reward function, thereby establishing an exact relationship between GFNs and maximum entropy …

abstract alternative arxiv cs.lg entropy flow generative inspiration markov maximum mcmc networks objects q-learning reinforcement reinforcement learning sampling scalable tool type while