reproducibilityindex.ai

Invariant Causal Prediction for Block MDPs

Authors: Amy Zhang, Clare Lyle, Shagun Sodhani, Angelos Filos, Marta Kwiatkowska, Joelle Pineau, Yarin Gal, Doina Precup

Reproducibility Variable	Result	LLM Response
Research Type	Experimental	We give empirical evidence that our methods work in both linear and nonlinear settings, attaining improved generalization over singleand multi-task baselines.
Researcher Affiliation	Collaboration	1Mc Gill University 2Mila 3Facebook AI Research 4University of Oxford 5Deepmind.
Pseudocode	Yes	Algorithm 1 Linear MISA Algorithm 2 Nonlinear Model-irrelevance State Abstraction (MISA) Learning
Open Source Code	Yes	Code is available at https://github.com/facebookresearch/ icp-block-mdp.
Open Datasets	Yes	We randomly initialize the background color of two train environments from Deepmind Control (Tassa et al., 2018) from range [0, 255].
Dataset Splits	No	The paper describes training and test environments but does not explicitly mention a separate validation set or how data was split for validation purposes.
Hardware Specification	No	The paper does not provide specific details about the hardware used for experiments, such as GPU models, CPU models, or memory specifications.
Software Dependencies	No	The paper does not specify the versions of any software dependencies (e.g., libraries, frameworks) used in the experiments.
Experiment Setup	Yes	Implementation details found in Appendix C.1. Implementation details and more information about Soft Actor Critic can be found in Appendix C.2. Additional plots with the hyperparameter sweep done to ﬁnd a good penalty weight for IRM can also be found Appendix C.3.