GFlowNets are probabilistic models that sequentially generate compositional structures through a stochastic policy. Among GFlowNets, temperature-conditional GFlowNets can introduce temperature-based controllability for exploration and exploitation. We propose \textit{Logit-scaling GFlowNets} (Logit-GFN), a novel architectural design that greatly accelerates the training of temperature-conditional GFlowNets. It is based on the idea that previously proposed approaches introduced numerical challenges in the deep network training, since different temperatures may give rise to very different gradient profiles as well as magnitudes of the policy's logits. We find that the challenge is greatly reduced if a learned function of the temperature is used to scale the policy's logits directly. Also, using Logit-GFN, GFlowNets can be improved by having better generalization capabilities in offline learning and mode discovery capabilities in online learning, which is empirically verified in various biological and chemical tasks. Our code is available at \url{https://github.com/dbsxodud-11/logit-gfn}

翻译：GFlowNet是一种通过随机策略顺序生成组合结构的概率模型。其中，温度条件GFlowNet能够引入基于温度的可控性，以平衡探索与利用。我们提出*对数几率缩放GFlowNet*（Logit-GFN），这是一种新颖的架构设计，可大幅加速温度条件GFlowNet的训练。其核心思想在于：先前提出的方法在深度网络训练中引入了数值挑战，因为不同温度可能导致策略的对数几率产生截然不同的梯度分布及量级。我们发现，若使用温度的习得函数直接缩放策略的对数几率，该挑战将显著缓解。此外，采用Logit-GFN能够增强GFlowNet在离线学习中的泛化能力及在线学习中的模式发现能力，这一结论已在多种生物学与化学任务中得到经验验证。我们的代码开源在\url{https://github.com/dbsxodud-11/logit-gfn}。