Curriculum Design for Teaching via Demonstrations: Theory and Applications

Authors: Gaurav Yengera, Rati Devidze, Parameswaran Kamalaruban, Adish Singla

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on a synthetic car driving environment and navigation-based environments demonstrate the effectiveness of our curriculum strategy.
Researcher Affiliation Academia 1Max Planck Institute for Software Systems (MPI-SWS), Saarbrucken, Germany 2Saarland University, Saarland Informatics Campus (SIC), Saarbrucken, Germany 3The Alan Turing Institute, London, UK
Pseudocode Yes Algorithm 1 Teacher-Learner Interaction 1: Initialization: Initial knowledge of learner πL 1 . 2: for t = 1, 2, . . . do 3: Teacher observes the learner s current policy πL t . 4: Teacher provides demonstration ξt to the learner. 5: Learner updates its policy to πL t+1 using ξt.
Open Source Code Yes 1Github repo: https://github.com/adishs/neurips2021_curriculum-teaching-demonstrations_code.
Open Datasets No The paper describes synthetic environments ("synthetic car driving environment", "navigation-based environments") and how they are constructed, but does not refer to a pre-existing, publicly available dataset with a formal citation, link, or repository information.
Dataset Splits No The paper mentions "training" and "test sets" but does not specify explicit numerical splits (e.g., percentages or counts) for training, validation, or test data. It also does not cite a standard split for any dataset used.
Hardware Specification No The paper states that hardware details are provided in the Appendix, but the Appendix content is not included in the provided text. Therefore, no specific hardware models, processors, or compute resources are detailed within the scope of the provided document.
Software Dependencies No The paper mentions that training details are in the Appendix and code, but does not provide specific software dependencies with version numbers (e.g., Python 3.x, PyTorch 1.x) within the provided text.
Experiment Setup No The paper describes the general setup, such as using a "6-layer Convolutional Neural Network (CNN)" and