Improving Feasibility via Fast Autoencoder-based Projections
Maria Chzhen, Priya L. Donti · 2026
Read Improving Feasibility via Fast Autoencoder-based Projections by Maria Chzhen, Priya L. Donti (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-2604-03489.txt.
Description
Chzhen and Donti propose a fast autoencoder-based fixer that fixes rule-breaking guesses in learning and control systems with complex working rules. A trained autoencoder learns a structured, convex latent picture of the feasible set, then projects latent codes onto a simple convex shape before reading back into the original feasible set. Tests on constrained tuning and reward-based learning problems show solid rule upkeep at low compute cost.
Catalog record
| CREATOR | Maria Chzhen, Priya L. Donti |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70683 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.03489