From Garbage to Gold: A Data-Architectural Theory of Predictive Robustness
Terrence J. Lee-St. John [email protected], Jordan L. Lawson [email protected], Bartlomiej Piechowski-Jozwiak [email protected] · 2026
Read From Garbage to Gold: A Data-Architectural Theory of Predictive Robustness by Terrence J. Lee-St. John [email protected], Jordan L. Lawson [email protected], Bartlomiej Piechowski-Jozwiak [email protected] (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-2603-12288.txt.
Introduction
Lee-St John, Lawson, and Piechowski-Jozwiak explain why messy high-dimensional data can still predict well. Modern tabular models thrive on collinear, error-prone inputs, against the garbage-in, garbage-out mantra. Their theory joins information theory, latent factor models, and psychometrics in one account. Predictor noise splits into predictor error and structural uncertainty from stochastic mappings. Large sets of flawed predictors beat both noise types in the limit, while cleaned small sets stay bounded. Informative collinearity helps too, since shared latent causes raise reliability and ease latent inference.
Catalog record
| CREATOR | Terrence J. Lee-St. John [email protected], Jordan L. Lawson [email protected], Bartlomiej Piechowski-Jozwiak [email protected] |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 325706 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.12288