Error estimates for tamed Euler and Randomized Euler schemes for SDEs with locally Lipschitz drift with applications to non-logconcave sampling and optimization
Iosif Lytras †, Angelos Ntousis ∗ · 2026
Read Error estimates for tamed Euler and Randomized Euler schemes for SDEs with locally Lipschitz drift with applications to non-logconcave sampling and optimization by Iosif Lytras †, Angelos Ntousis ∗ (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-2605-24937.txt.
Description
Iosif Lytras and Angelos Ntousis analyze schemes for stochastic differential equations with super-linear, locally Lipschitz drift, where the Euler-Maruyama method is unstable. They study the KL-accelerated tamed unadjusted Langevin algorithm and add a tamed randomized midpoint method. Two frameworks give non-asymptotic estimates: KL divergence for one scheme and total variation for the other. Under a logarithmic Sobolev inequality, the first scheme needs near-optimal epsilon to the minus one half KL iterations. The second gives a total variation guarantee under super-linear drift.
Catalog record
| CREATOR | Iosif Lytras †, Angelos Ntousis ∗ |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 140438 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.24937