Robust Communicative Multi-Agent Reinforcement Learning with Active Defense

Authors: Lebin Yu, Yunbo Qiu, Quanming Yao, Yuan Shen, Xudong Zhang, Jian Wang

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

Reproducibility Variable Result LLM Response
Research Type Experimental The superiority of ADMAC over existing methods is validated by experiments in three communication-critical tasks under four types of attacks.
Researcher Affiliation Academia Department of Electronic Engineering, BNRist, Tsinghua University, Beijing 100084, China
Pseudocode No The provided paper does not contain explicit pseudocode or algorithm blocks in its main content.
Open Source Code No The paper does not contain an explicit statement or link confirming the public availability of the source code for the methodology described.
Open Datasets Yes We implement three communication-critical multi-agent environments for demonstrative purposes: Food Collector (Sun et al. 2023), Predator Prey (Singh, Jain, and Sukhbaatar 2018), and Treasure Hunt (Freed et al. 2020).
Dataset Splits No The paper describes how the dataset for the reliability estimator is generated, but it does not specify explicit training, validation, and test dataset splits with percentages or counts for reproducibility.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers required to replicate the experiments.
Experiment Setup Yes We set σ = 0.5 in the experiments. ... we set η = 1 to obtain a strong attack. ... We set ϵ = 0.3 and use 5-step updates to obtain the final perturbed messages.