TY  - RPRT
TI  - A recipe for scalable attention-based MLIPs: unlocking long-range accuracy with all-to-all node attention
AU  - Eric Qu
AU  - Brandon M. Wood
AU  - Aditi S. Krishnapriyan
AU  - Zachary W. Ulissi
PY  - 2026
UR  - https://arxiv.org/abs/2603.06567
ID  - 2603.06567
ER  -