Policy Teaching via Environment Poisoning: Training-time Adversarial Attacks against Reinforcement Learning

Authors: Amin Rakhsha, Goran Radanovic, Rati Devidze, Xiaojin Zhu, Adish Singla

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

Reproducibility Variable Result LLM Response
Research Type Experimental We perform numerical simulations on an environment represented as an MDP with four states and two actions, see Figure 2 for details.
Researcher Affiliation Academia 1Max Planck Institute for Software Systems (MPI-SWS). 2University of Wisconsin-Madison.
Pseudocode No The paper describes its methods through mathematical formulations and prose, but does not include any explicit pseudocode or algorithm blocks.
Open Source Code Yes The implementation details and code are provided in supplementary materials.
Open Datasets No The environment is custom-defined and described in Figure 2 and Section 6, but no public access information (link, citation) is provided for it as a dataset.
Dataset Splits No The paper conducts simulations within a defined MDP environment and does not describe dataset splits for training, validation, or testing.
Hardware Specification No The paper does not specify any hardware details (e.g., CPU, GPU models, or memory) used for running the experiments.
Software Dependencies No The paper mentions using the UCRL learning algorithm but does not specify any software libraries or packages with version numbers used for implementation.
Experiment Setup Yes The regularity parameter δ in the problems for solving dynamic poisoning attacks is set to 0.0001. In the experiments, we fix R(s0, .) = 2.5 and ϵ = 0.1