NeuroPareto: Calibrated Acquisition for Costly Many-Goal Search in Vast Parameter Spaces
Rong Fu, Chunlei Meng, Youjin Wang, Haoyu Zhao, Jiaxuan Lu, Kun Liu, JiaBao Dou, Simon James Fong · 2026
Read NeuroPareto: Calibrated Acquisition for Costly Many-Goal Search in Vast Parameter Spaces by Rong Fu, Chunlei Meng, Youjin Wang, Haoyu Zhao, Jiaxuan Lu, Kun Liu, JiaBao Dou, Simon James Fong (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-2602-03901.txt.
Description
Searching for the best trade-offs among many goals is expensive when each test costs real computation. Rong Fu and 6 co-authors present NeuroPareto, a 2026 system that hunts Pareto-optimal designs, meaning designs where no goal improves without hurting another. It targets vast parameter spaces with strict evaluation budgets.
A Bayesian classifier ranks candidates by quality tier and estimates uncertainty. Gaussian process models split that uncertainty into parts that testing can and cannot reduce. A small network, trained on past gains, picks which candidates deserve full evaluation. On standard suites and an energy extraction task, NeuroPareto beat comparable methods.
Catalog record
| CREATOR | Rong Fu, Chunlei Meng, Youjin Wang, Haoyu Zhao, Jiaxuan Lu, Kun Liu, JiaBao Dou, Simon James Fong |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 129 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-03901