Optimally Deceiving a Learning Leader in Stackelberg Games

Authors: Georgios Birmpas, Jiarui Gan, Alexandros Hollender, Francisco Marmolejo, Ninad Rajgopal, Alexandros Voudouris

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this paper, we fill this gap by showing that it is always possible for the follower to efficiently compute (near-)optimal payoffs for various scenarios of learning interaction between the leader and the follower. By exploiting an intuitive characterization of all strategy profiles that can be induced as SSEs in Stackelberg games, we show that it is always possible (irrespective of the learning algorithm employed by the leader) for the follower to compute, in polynomial time, a payoff matrix implying an SSE which maximizes his true utility. Furthermore, we strengthen this result to resolve possible equilibrium selection issues, by showing that the follower can construct a payoff matrix that induces a unique SSE, in which his utility is maximized up to some arbitrarily small loss.
Researcher Affiliation Academia Georgios Birmpas Sapienza University of Rome gebirbas@gmail.com Jiarui Gan University of Oxford jiarui.gan@cs.ox.ac.uk Alexandros Hollender University of Oxford alexandros.hollender@cs.ox.ac.uk Francisco J. Marmolejo-Cossío University of Oxford francisco.marmolejo@cs.ox.ac.uk Ninad Rajgopal University of Oxford ninad.rajgopal@cs.ox.ac.uk Alexandros A. Voudouris University of Essex alexandros.voudouris@essex.ac.uk
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not include any statement or link indicating that open-source code for the described methodology is available.
Open Datasets No This is a theoretical paper. It does not mention or use training datasets for empirical evaluation.
Dataset Splits No This is a theoretical paper. It does not mention or use validation sets for empirical evaluation.
Hardware Specification No This is a theoretical paper and does not mention any hardware specifications used for experiments.
Software Dependencies No This is a theoretical paper and does not list specific software dependencies with version numbers for reproducibility.
Experiment Setup No This is a theoretical paper and does not describe an experimental setup with specific hyperparameters or system-level training settings.