Learning Portable Representations for High-Level Planning

Authors: Steven James, Benjamin Rosman, George Konidaris

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

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate transfer in two different domains, where an agent learns a portable, task-independent symbolic vocabulary, as well as operators expressed in that vocabulary, and then learns to instantiate those operators on a per-task basis. This reduces the number of samples required to learn a representation of a new task.
Researcher Affiliation Academia 1School of Computer Science and Applied Mathematics, University of the Witwatersrand, Johannesburg, South Africa 2Department of Computer Science, Brown University, Providence RI 02912, USA.
Pseudocode No The paper describes processes using flowcharts and text but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any information about open-sourcing its code.
Open Datasets No The paper introduces two custom domains, 'Rod-and-Block' and 'Treasure Game', stating 'We construct a domain we term Rod-and-Block' and 'We construct a set of ten tasks... of the Treasure Game'. No public access information (link, citation, repository) is provided for these datasets.
Dataset Splits No The paper describes how samples are collected and used for model building and evaluation but does not specify explicit train/validation/test splits (e.g., percentages or exact counts for data partitioning).
Hardware Specification No The paper does not provide any specific hardware details (e.g., GPU/CPU models, memory, cloud resources) used for running the experiments.
Software Dependencies No The paper mentions using DBSCAN, Support Vector Machine with Platt scaling, and Kernel Density Estimation, but does not provide specific version numbers for any of these software components or libraries.
Experiment Setup No The paper describes the general experimental procedure (e.g., collecting k transition samples, using a threshold of 0.75 for accuracy) but does not provide specific hyperparameter values or detailed system-level training settings for the models (e.g., learning rates, batch sizes, optimizer configurations).