SURF: Steering the Scalarization Weight to Uniformly Traverse the Pareto Front
Liuyuan Jiang, Chentong Huang, Lisha Chen · 2026
Read SURF: Steering the Scalarization Weight to Uniformly Traverse the Pareto Front by Liuyuan Jiang, Chentong Huang, Lisha Chen (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-2605-20619.txt.
Introduction
Liuyuan Jiang, Chentong Huang, and Lisha Chen propose SURF, a way to sample uniformly along the Pareto front in multi-objective optimization. Evenly spaced scalarization weights usually give uneven front coverage, since solutions traverse the front at nonuniform speed as weights vary. Integrating that speed gives cumulative arc length, and normalizing it yields a distribution function on weights. Its inverse picks points at equal arc-length gaps, and the authors derive closed forms for structured cases like bi-objective bandits.
Catalog record
| CREATOR | Liuyuan Jiang, Chentong Huang, Lisha Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 207986 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.20619