Last-Iterate Convergence for Generalized Frank-Wolfe in Monotone Variational Inequalities

Authors: Zaiwei Chen, Eric Mazumdar

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we conduct numerical simulations to empirically verify the performance of our proposed algorithms.
Researcher Affiliation Academia Zaiwei Chen Purdue IE West Lafayette, IN 47907 chen5252@purdue.edu Eric Mazumdar Caltech CMS Pasadena, CA 91125 mazumdar@caltech.edu
Pseudocode Yes Algorithm 1 Smoothed Fictitious Play (from Player 1 s perspective); Algorithm 2 Generalized Frank-Wolfe for Monotone Variational Inequalities; Algorithm 3 Stochastic Frank-Wolfe for Monotone Variational Inequalities
Open Source Code No This paper is a theoretical work, and the numerical simulations are conducted on synthetic examples to demonstrate the effectiveness of the proposed algorithm.
Open Datasets No The paper uses synthetic game setups (Rock-Paper-Scissors, Burglar-Policeman, Randomly Generated Matrix Game) for numerical simulations, but does not provide concrete access information or formal citations for these as publicly available datasets.
Dataset Splits No The numerical simulations describe game setups and parameter choices but do not specify explicit training, validation, or test splits for the data.
Hardware Specification No The paper states that it is theoretical work with numerical simulations on synthetic examples, and thus does not provide specific hardware details (like GPU/CPU models or memory) used for running experiments.
Software Dependencies No The paper does not provide specific version numbers for any software dependencies used in the numerical simulations.
Experiment Setup Yes In the stochastic setting, we choose τ = 0.1, α = 0.1, and β = 0.01 in Algorithm 3.