Saved in:
Bibliographic Details
Main Authors: Kim, Hohyun, Lee, Seunggeun, Oh, Min-hwan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2506.02685
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918161409900544
author Kim, Hohyun
Lee, Seunggeun
Oh, Min-hwan
author_facet Kim, Hohyun
Lee, Seunggeun
Oh, Min-hwan
contents Generative Flow Networks (GFlowNets) offer a powerful framework for sampling graphs in proportion to their rewards. However, existing approaches suffer from systematic biases due to inaccuracies in state transition probability computations. These biases, rooted in the inherent symmetries of graphs, impact both atom-based and fragment-based generation schemes. To address this challenge, we introduce Symmetry-Aware GFlowNets (SA-GFN), a method that incorporates symmetry corrections into the learning process through reward scaling. By integrating bias correction directly into the reward structure, SA-GFN eliminates the need for explicit state transition computations. Empirical results show that SA-GFN enables unbiased sampling while enhancing diversity and consistently generating high-reward graphs that closely match the target distribution.
format Preprint
id arxiv_https___arxiv_org_abs_2506_02685
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Symmetry-Aware GFlowNets
Kim, Hohyun
Lee, Seunggeun
Oh, Min-hwan
Machine Learning
Generative Flow Networks (GFlowNets) offer a powerful framework for sampling graphs in proportion to their rewards. However, existing approaches suffer from systematic biases due to inaccuracies in state transition probability computations. These biases, rooted in the inherent symmetries of graphs, impact both atom-based and fragment-based generation schemes. To address this challenge, we introduce Symmetry-Aware GFlowNets (SA-GFN), a method that incorporates symmetry corrections into the learning process through reward scaling. By integrating bias correction directly into the reward structure, SA-GFN eliminates the need for explicit state transition computations. Empirical results show that SA-GFN enables unbiased sampling while enhancing diversity and consistently generating high-reward graphs that closely match the target distribution.
title Symmetry-Aware GFlowNets
topic Machine Learning
url https://arxiv.org/abs/2506.02685