Exploiting Multiple Abstractions in Episodic RL via Reward Shaping

Authors: Roberto Cipollone, Giuseppe De Giacomo, Marco Favorito, Luca Iocchi, Fabio Patrizi

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Moreover, we prove that the method guarantees optimal convergence and we demonstrate its effectiveness experimentally. and (iv) an experimental analysis showing that our approach significantly improves sample-efficiency and that modelling errors yield only a limited performance degradation.
Researcher Affiliation Collaboration Roberto Cipollone1, Giuseppe De Giacomo1,2, Marco Favorito3, Luca Iocchi1, Fabio Patrizi1 1DIAG, Universit a degli Studi di Roma La Sapienza , Italy 2Department of Computer Science, University of Oxford, U.K. 3Banca d Italia, Italy
Pseudocode Yes Algorithm 1: Main algorithm
Open Source Code Yes Code available at https://github.com/cipollone/multinav2
Open Datasets No The paper describes custom-built environments like the '4-rooms' and '8-rooms' domain. While the code for the environment might be implied to be in the linked GitHub repo, the paper does not provide concrete access information (link, DOI, formal citation with authors/year, or reference to established benchmark datasets) for a publicly available or open dataset as a standalone dataset.
Dataset Splits No The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions the use of algorithms like Q-learning, Delayed Q-learning, and Dueling DQN, but does not provide specific ancillary software details, such as library or solver names with version numbers (e.g., Python version, PyTorch/TensorFlow versions).
Experiment Setup No The paper mentions 'Further training details can be found in the appendix' but does not include specific experimental setup details (concrete hyperparameter values, training configurations, or system-level settings) in the main text.