Classical and Quantum Speedups for Non-Convex Optimization via Energy Conserving Descent

Yihang Sun, Huaijin Wang, Patrick Hayden, Jose Blanchet · 2026

Read Classical and Quantum Speedups for Non-Convex Optimization via Energy Conserving Descent by Yihang Sun, Huaijin Wang, Patrick Hayden, Jose Blanchet (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-2604-13022.txt.

Description

Sun, Wang, Hayden, and Blanchet give the first analytical study of energy-conserving descent, an optimization method that keeps energy fixed instead of draining it away. For one-dimensional double-well objectives, they compute hitting times from a local minimum to the global one. They prove exponential speedups over gradient baselines in both stochastic and quantum versions. The quantum version gains further on tall barriers, since energy conservation sidesteps the rare escapes that slow stochastic gradient descent.

Catalog record

CREATORYihang Sun, Huaijin Wang, Patrick Hayden, Jose Blanchet
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74943
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.13022

Browse the full library · About the Tower of Babel