ICML

Adversarial Robustness of Graph Transformers

图变换器的对抗鲁棒性

Philipp Foth, Lukas Gosch, Simon Geisler, Leo Schwinn, Stephan Günnemann

机构 * Munich Center for Machine Learning (MCML)（慕尼黑机器学习中心） 纠错

AI总结 研究图变换器在结构扰动下的对抗鲁棒性，设计了首个自适应攻击方法，评估了多种任务和扰动模型，发现图变换器在许多情况下存在严重脆弱性。

Comments TMLR 2025 (J2C-Certification: Presented @ ICLR 2026). A preliminary version appeared at the Differentiable Almost Everything Workshop at ICML 2024. Code available at https://github.com/isefos/gt_robustness

Non-stationary Diffusion For Probabilistic Time Series Forecasting

非平稳扩散用于概率时间序列预测

Weiwei Ye, Zhuopeng Xu, Ning Gui

机构 * Central South University（中南大学） 纠错

AI总结 本文提出非平稳扩散模型，通过改进噪声模型以适应时间变化的不确定性，提升时间序列预测性能。

Comments Accepted as spotlight poster at ICML