Using ”AI Poincaré” to analyze non-linear integrable optics
Lazare Osmanov, Nilanjan Banerjee · 2025
Read Using ”AI Poincaré” to analyze non-linear integrable optics by Lazare Osmanov, Nilanjan Banerjee (2025) 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-2510-22705.txt.
Description
Some values stay fixed while a system moves. Finding those values helps in the study of accelerator optics. Lazare Osmanov and Nilanjan Banerjee wrote a 2025 paper on a tool for that hunt. The tool is called AI Poincare. It finds conserved quantities on its own from data.
Their test case is the McMillan system of optics. They score the tool on its count of fixed values. They also check how close its inferred surface lies to true data points. A best range of perturbation size is found. A new neural network design is proposed from the results. Data from a test accelerator at Fermilab were used too. The tool found the right count even with fast signal decay.
Catalog record
| CREATOR | Lazare Osmanov, Nilanjan Banerjee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-22705