paper-with-me

홈 › Papers

Flatness-Aware Stochastic Gradient Langevin Dynamics

2025-10-02 · Stefano Bruno, Youngsik Hwang, Jaehyeon An, Sotirios Sabanis, Dong-Young Lim arxiv

Flatness of the loss landscape has been widely studied as an important perspective for understanding the behavior and generalization of deep learning algorithms. Motivated by this view, we propose Flatness-Aware Stochastic Gradient Langevin Dynamics (fSGLD), a first-order optimization method that biases learning its dynamics toward flat basins while retaining the computational and memory efficiency of SGD and SGLD. We provide a non-asymptotic theoretical analysis showing that fSGLD targets a flatness-biased Gibbs distribution under a theoretically prescribed coupling between the noise scale $σ$ and the inverse temperature $β$, together with explicit excess risk guarantees. We empirically evaluate fSGLD across standard optimizer benchmarks, Bayesian image classification, uncertainty quantification, and out-of-distribution detection, demonstrating consistently strong performance and reliable uncertainty estimates. Additional experiments confirm the effectiveness of the theoretically prescribed $β$-$σ$ coupling compared to decoupled choices.

📄 PDF Abstract BibTeX arXiv:2510.02174

Code (0)

등록된 구현이 없습니다.

Tasks

Out-of-Distribution DetectionImage Classification

Similar Papers 제목 키워드 기반

Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization

2017-07-20 · NeurIPS 2018 12 · Pan Xu, Jinghui Chen, Difan Zou, Quanquan Gu

We present a unified framework to analyze the global convergence of Langevin dynamics based algorithms for nonconvex finite-sum optimization with $n$ component functions. At the core of our analysis is a direct analysis …

Variance Reduction in Stochastic Gradient Langevin Dynamics

2016-12-01 · NeurIPS 2016 12 · Kumar Avinava Dubey, Sashank J. Reddi, Sinead A. Williamson, Barnabas Poczos 외

Stochastic gradient-based Monte Carlo methods such as stochastic gradient Langevin dynamics are useful tools for posterior inference on large scale datasets in many machine learning applications. These methods scale to l…

BIG-bench Machine Learning

Federated Stochastic Gradient Langevin Dynamics

2020-04-23 · Khaoula El Mekkaoui, Diego Mesquita, Paul Blomstedt, Samuel Kaski

Stochastic gradient MCMC methods, such as stochastic gradient Langevin dynamics (SGLD), employ fast but noisy gradient estimates to enable large-scale posterior sampling. Although we can easily extend SGLD to distributed…

Federated LearningMetric Learning

Non-Convex Optimization via Non-Reversible Stochastic Gradient Langevin Dynamics

2020-04-06 · Yuanhan Hu, Xiaoyu Wang, Xuefeng Gao, Mert Gurbuzbalaban 외

Stochastic Gradient Langevin Dynamics (SGLD) is a powerful algorithm for optimizing a non-convex objective, where a controlled and properly scaled Gaussian noise is added to the stochastic gradients to steer the iterates…

Stochastic Optimization

Information-Theoretic Generalization Bounds for SGLD via Data-Dependent Estimates

2019-11-06 · NeurIPS 2019 12 · Jeffrey Negrea, Mahdi Haghifam, Gintare Karolina Dziugaite, Ashish Khisti 외

In this work, we improve upon the stepwise analysis of noisy iterative learning algorithms initiated by Pensia, Jog, and Loh (2018) and recently extended by Bu, Zou, and Veeravalli (2019). Our main contributions are sign…

Generalization Bounds