Understanding Generalization in Physics Informed Models through Affine Variety Dimensions
Takeshi Koshizuka, Issei Sato · 2025
Read Understanding Generalization in Physics Informed Models through Affine Variety Dimensions by Takeshi Koshizuka, Issei Sato (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-2501-18879.txt.
Description
Takeshi Koshizuka and Issei Sato present a 2025 theory paper on physics-informed learning. Such models blend known physics with real data to learn faster from fewer samples. Past theory leaned on full prior knowledge and straight-line systems, while real problems mix partial rules with curved behavior.
The authors give one shared form that covers the main ways physics is folded into training. In that frame, how well the model generalizes turns on the size of a shape tied to the physics rule, not on the count of model settings. They also give a way to gauge that size, and tests back the claim.
Catalog record
| CREATOR | Takeshi Koshizuka, Issei Sato |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 109 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-18879