English(EN) Self-Supervised Pretraining of Molecular Graph Encoders with LeJEPA

LeJEPA 架构已适配用于分子图编码器

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-07 04:00

研究人员已将自监督预训练架构 LeJEPA 应用于分子图神经网络，以评估其对分子性质预测的影响。预训练虽然增强了学习到的表示，但并未持续提升在抗生素活性预测或分子性质预测（ogbg-molhiv）等任务上的微调性能。然而，当预训练嵌入作为冻结探针使用时，其性能显著优于随机初始化，并且将这些嵌入与传统的 Morgan 指纹相结合，在 ogbg-molhiv 基准测试中取得了最佳性能。 AI

影响该研究探索了使用自监督学习改进分子性质预测的方法，有望推动药物发现和材料科学的发展。

排序理由学术论文，详细介绍了一种新的分子图编码器方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 32 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 学术论文，详细介绍了一种新的分子图编码器方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Micha{\l} Kulczykowski, Rafa{\l} {\L}ab\k{e}dzki · 2026-09-07 04:00

Self-Supervised Pretraining of Molecular Graph Encoders with LeJEPA

arXiv:2609.04261v1 Announce Type: cross Abstract: Self-supervised pretraining has transformed language and vision, but its value for molecular graph neural networks remains contested. We ask whether pretraining on a large unlabelled corpus improves molecular property prediction. …

本文如何被排名

报道来源 [1]

Self-Supervised Pretraining of Molecular Graph Encoders with LeJEPA

相关实体

相关话题