Learning Task Specifications from Demonstrations

Authors: Marcell Vazquez-Chanlatte, Susmit Jha, Ashish Tiwari, Mark K. Ho, Sanjit Seshia

NeurIPS 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In our experiments, we demonstrate how learning specifications can help avoid common problems that often arise due to ad-hoc reward composition.Running a fairly unoptimized implementation of Algorithm 2 on the concept class and demonstrations took approximately 95 seconds and resulted in 172 eϕ queries ( 18% of the concept class).
Researcher Affiliation Collaboration 1 University of California, Berkeley 2 SRI International, Menlo Park
Pseudocode Yes Algorithm 1 Inference on chains and Algorithm 2 Inference on Partial Orders
Open Source Code No No explicit statement or link providing access to the source code for the methodology described in this paper was found.
Open Datasets No The paper uses custom demonstrations shown in Figure 1 for its experiments but does not provide concrete access information (link, DOI, or formal citation) for a publicly available or open dataset.
Dataset Splits No The paper uses a set of demonstrations for inference but does not specify any explicit training, validation, or test dataset splits.
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, memory) used for running experiments are provided in the paper.
Software Dependencies No The paper mentions using a "SAT solver" and "Binary Decision Diagram (BDD)" but does not specify version numbers for these or any other software dependencies.
Experiment Setup No The paper describes the scenario and concept class but does not provide specific experimental setup details such as hyperparameters, training configurations, or other system-level settings.