Improved Sample Complexity for Incremental Autonomous Exploration in MDPs

Authors: Jean Tarbouriech, Matteo Pirotta, Michal Valko, Alessandro Lazaric

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

Reproducibility Variable Result LLM Response
Research Type Experimental Finally, we report preliminary empirical results confirming our theoretical findings. [...] We observe that Dis Co outperforms Ucb Explore for any value of .
Researcher Affiliation Collaboration Jean Tarbouriech Facebook AI Research Paris & Inria Lille Matteo Pirotta Facebook AI Research Paris Michal Valko Deep Mind Paris Alessandro Lazaric Facebook AI Research Paris
Pseudocode Yes Algorithm 1: Algorithm Dis Co
Open Source Code No The paper does not contain any explicit statements or links indicating that the source code for their methodology is publicly available.
Open Datasets No The paper mentions environments like "confusing chain domain" and "combination lock problem introduced in [31]" for numerical simulations, but does not provide concrete access information (e.g., URL, DOI, repository, or formal citation for data) for these as publicly available datasets.
Dataset Splits No The paper mentions that "Values are averaged over 50 runs" but does not provide specific training, validation, or test dataset splits. The experimental setup uses simulated environments that generate data through interaction, not pre-defined static datasets with splits.
Hardware Specification No The paper does not provide any specific details about the hardware used to run the experiments (e.g., GPU/CPU models, processors, memory).
Software Dependencies No The paper does not provide specific version numbers for any software dependencies. It only mentions general aspects like removing "logarithmic and constant terms for simplicity" and using "empirical Bernstein inequality".
Experiment Setup No The main text states: "We refer the reader to App. F for details on the algorithmic configurations and on the environments considered." This indicates that specific experimental setup details are not provided in the main body of the paper.