Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

Fisher information dissipation for time-inhomogeneous stochastic differential equations

Authors: Qi Feng, Xinzhe Zuo, Wuchen Li

JMLR 2024 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Numerical examples demonstrate the convergence results for the time-dependent Langevin dynamics. Several numerical experiments are provided to justify our theoretical results.
Researcher Affiliation Academia Qi Feng EMAIL Department of Mathematics Florida State University Tallahassee, FL 32306, USA; Xinzhe Zuo EMAIL Department of Mathematics University of California, Los Angeles Los Angeles, CA 90095, USA; Wuchen Li EMAIL Department of Mathematics University of South Carolina Columbia, SC 29208, USA
Pseudocode No The paper describes mathematical formulations and derivations, along with numerical schemes like Euler-Maruyama in equations, but does not present any clearly labeled pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any statement or link regarding the availability of source code for the methodology described.
Open Datasets No The paper's numerical experiments use synthetic data. For example, it states, "We first sample M = 10^6 particles from N(0, 1)" and simulates stochastic differential equations, rather than using external publicly available datasets.
Dataset Splits No The numerical experiments are based on simulations rather than external datasets with predefined splits. There are no mentions of training, validation, or test splits.
Hardware Specification No The paper does not specify any particular hardware (e.g., GPU, CPU models) used for running the numerical experiments.
Software Dependencies No The paper mentions using the "Euler-Maruyama scheme" for discretization but does not specify any software libraries, frameworks, or their version numbers.
Experiment Setup Yes We first sample M = 10^6 particles from N(0, 1). Then we evolve (37) using the Euler-Maruyama scheme shown below for N = 10000 steps with a step size of h = 0.002: Xn+1 = Xn - h∇V (Xn) + √(2C log(nh + t0))Bn , where Bn ∼ N(0, √h), and t0 = e. [...] we choose K = 50 in our numerical experiment. [...] In two-dimension (Fig. 3), we used M = 10^6 particles, N = 10000 steps with a stepsize of h = 0.001.