That’s Mine! Learning Ownership Relations and Norms for Robots
Authors: Zhi-Xuan Tan, Jake Brawer, Brian Scassellati8058-8065
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Through a series of simulated and real-world experiments, we demonstrate the competence and flexibility of the system in performing object manipulation tasks that require a variety of norms to be followed, laying the groundwork for future research into the acquisition and application of social norms. |
| Researcher Affiliation | Academia | Zhi-Xuan Tan,1,2 Jake Brawer,1 Brian Scassellati1 1Department of Computer Science, Yale University, New Haven, CT, USA 2A*STAR Artificial Intelligence Initiative, Agency for Science, Technology and Research (A*STAR), Singapore |
| Pseudocode | Yes | Algorithm 1 Cover a positive example; Algorithm 2 Uncover a negative example; Algorithm 3 Add a rule after refinement; Algorithm 4 Subtract a rule after refinement |
| Open Source Code | No | The paper provides a link to a video demonstration: 'To demonstrate the system s capabilities in the real world, we provide a video at https://bit.ly/2z8obET.', but it does not provide concrete access to the source code for the methodology described. |
| Open Datasets | No | The paper describes a 'simulated environment of 20 colored blocks and 3 agents, with 5 blocks unowned and 5 blocks owned by each agent' and details how this data was generated. It does not provide concrete access information (link, DOI, repository, or formal citation for public access) for this custom simulated dataset. |
| Dataset Splits | No | The paper mentions 'sequentially providing both the true ownership relations and objectspecific permissions for half of the objects at random, and then evaluating ownership accuracy for the other half' and evaluating 'accuracy and F1 measure of the induced rules' for different percentages of provided permissions. While this describes data usage, it does not explicitly define or specify distinct training, validation, and test splits using those terms with exact percentages, sample counts, or citations to predefined splits typically required for reproducibility of data partitioning. |
| Hardware Specification | No | The paper mentions deployment on the 'Baxter robotic platform' for real-world demonstration but does not provide specific details about the computational hardware (e.g., GPU/CPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions 'Prob Log' and 'kernel logistic regression (KLR)' as components but does not provide specific version numbers for any software dependencies. |
| Experiment Setup | No | The paper describes elements of the simulated environment and some parameters (e.g., 'rate parameter 0.1 (1 interaction every 10 seconds)' for data generation, 'user-specified threshold for rule obedience'), but it does not provide specific machine learning hyperparameters (like learning rate, batch size, number of epochs for KLR or rule induction) or detailed system-level training configurations typically found in an experimental setup section. |