Learning missing physics from legacy simulators with alternating neural integrators
Hao Wang, Qinghe Wang, Caiyou Yuan, Kailiang Wu · 2026
Read Learning missing physics from legacy simulators with alternating neural integrators by Hao Wang, Qinghe Wang, Caiyou Yuan, Kailiang Wu (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/PMC13444323.txt.
Description
Hao Wang and colleagues patch old simulators with neural fixes that learn missing physics in a 2026 paper. Trusted simulators drift from truth where physics went unresolved, yet their code often stays closed. Their frame, Alternating Neural Integrators, swaps steps between the fixed old code and a learned correction.
The split follows operator-splitting rules, so each half does what it does best. Chaotic systems regained lost coupling under the patch in their tests. Turbulence runs used the patch as a fine-grid fix where old models wander. In neat settings, symbol distillation squeezed the patch into compact formulas that fed back into the old code. The authors pitch the frame as a gray-box upgrade path for labs stuck with legacy tools.
Catalog record
| CREATOR | Hao Wang, Qinghe Wang, Caiyou Yuan, Kailiang Wu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 94736 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |