DisParQ: Self-Supervised Part Concepts for Interpretable Vision Foundation Models

作者：Adam Pardyl, Siddhartha Gairola, Sukrut Rao, Adam Wróbel, Bartosz Zieliński, Bernt Schiele, Dawid Rymarczyk · 单位：† † thanks: * Equal contribution. $ˆ1$Jagiellonian University, Faculty of Mathematics and Computer Science, $ˆ2$Jagiellonian University, Doctoral School of Exact and Natural Sciences, $ˆ3$Max Planck Institute for Informatics, Saarland Informatics Campus, $ˆ4$Jagiellonian University, Jagiellonian Center for Artificial Intelligence · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-10-08