arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2406.13655cs.LGcs.AI
收藏

利用蒙特卡洛树搜索改进GFlowNets

Improving GFlowNets with Monte Carlo Tree Search

HSE University(高等经济大学);Skoltech(斯科尔科沃科技学院);CMAP – CNRS – École polytechnique – Institut Polytechnique de Paris(CMAP – 法国国家科学研究中心 – 巴黎综合理工学院 – 巴黎理工学院);Université Paris-Saclay(巴黎萨克雷大学);Steklov Mathematical Institute of Russian Academy of Sciences(俄罗斯科学院斯捷克洛夫数学研究所);Constructor University(康斯特鲁克大学)

Nikita Morozov, Daniil Tiapkin, Sergey Samsonov, Alexey Naumov, Dmitry Vetrov

AI总结:

该研究基于GFlowNets与熵正则化强化学习的关联,提出适配MENTS算法的蒙特卡洛树搜索方法以增强其规划能力,在训练和推理阶段应用后提升了训练样本效率与预训练模型生成保真度。

摘要:

生成流网络(Generative Flow Networks,GFlowNets)将组合离散空间上的分布采样视为序列决策问题,通过训练随机策略逐步构建对象。近期研究揭示了GFlowNets与熵正则化强化学习之间的紧密关联。基于这些发现,我们提出应用蒙特卡洛树搜索(Monte Carlo Tree Search,MCTS)来增强GFlowNets的规划能力。具体而言,我们展示了如何将MENTS算法(Xiao等人,2019)适配于GFlowNets,并在训练和推理阶段均加以使用。我们的实验表明,该方法提升了GFlowNet训练的样本效率,以及预训练GFlowNet模型的生成保真度。

英文摘要:

Generative Flow Networks (GFlowNets) treat sampling from distributions over compositional discrete spaces as a sequential decision-making problem, training a stochastic policy to construct objects step by step. Recent studies have revealed strong connections between GFlowNets and entropy-regularized reinforcement learning. Building on these insights, we propose to enhance planning capabilities of GFlowNets by applying Monte Carlo Tree Search (MCTS). Specifically, we show how the MENTS algorithm (Xiao et al., 2019) can be adapted for GFlowNets and used during both training and inference. Our experiments demonstrate that this approach improves the sample efficiency of GFlowNet training and the generation fidelity of pre-trained GFlowNet models.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑