Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger Conference: Proceedings of the ACM Web Conference 2026; April 13–17, 2026; Dubai, United Arab Emirates Proceedings of the ACM Web Conference 2026 (WWW ’26), April 13–17, 2026, Dubai, United Arab Emirates DOI: 10.1145/3774904.3792435 ISBN: 979-8-4007-2307-0/2026/04 CCS: Computing methodologies CCS: Information systems Web mining CCS: Computing methodologies Neural networks
Li Sun, Ming Zhang, Wenxin Jin, Zhongtian Sun, Zhenhao Huang, Hao Peng, Sen Su, Philip Yu · 2026
Read Heterophily-Agnostic Hypergraph Neural Networks with Riemannian Local Exchanger Conference: Proceedings of the ACM Web Conference 2026; April 13–17, 2026; Dubai, United Arab Emirates Proceedings of the ACM Web Conference 2026 (WWW ’26), April 13–17, 2026, Dubai, United Arab Emirates DOI: 10.1145/3774904.3792435 ISBN: 979-8-4007-2307-0/2026/04 CCS: Computing methodologies CCS: Information systems Web mining CCS: Computing methodologies Neural networks by Li Sun, Ming Zhang, Wenxin Jin, Zhongtian Sun, Zhenhao Huang, Hao Peng, Sen Su, Philip Yu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-00599.txt.
Description
Sun and seven colleagues make hypergraph neural nets work where linked nodes often differ and long range dependence matters. Standard message passing assumes similar neighbours and struggles at narrow choke points, a problem they link to oversquashing under Riemannian manifold heat flow. Their adaptive local heat exchanger tunes choke points per subgroup, capturing long range links through a Robin condition while preserving distinct representations via source terms. The design yields message passing that ignores heterophily, with theoretical guarantees.
Catalog record
| CREATOR | Li Sun, Ming Zhang, Wenxin Jin, Zhongtian Sun, Zhenhao Huang, Hao Peng, Sen Su, Philip Yu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67229 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.00599