Implementing Theory of Mind on a Robot Using Dynamic Epistemic Logic
Authors: Lasse Dissing, Thomas Bolander
IJCAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | This paper provides evidence for the claims by documenting the implementation of a DEL-based reasoning system on a humanoid robot. Our implementation allows the robot to perform cognitive perspective-taking, in particular to reason about the firstand higherorder beliefs of other agents. We demonstrate how this allows the robot to pass a quite general class of false-belief tasks involving human agents.In all 25 experiments, the robot correctly determined the false beliefs of participants |
| Researcher Affiliation | Academia | Lasse Dissing and Thomas Bolander DTU Compute, Denmark {ldiha, tobo}@dtu.dk |
| Pseudocode | No | The paper describes the Dynamic Epistemic Logic (DEL) framework and its application through formal definitions and examples, but it does not include pseudocode or algorithm blocks. |
| Open Source Code | No | The paper uses and cites external open-source libraries (e.g., Open Pose, dlib, Dan Speech), but it does not provide an explicit statement or link for the open-sourcing of the authors' own implementation code for the DEL-based reasoning system or the robot. |
| Open Datasets | No | The paper describes experiments conducted with 50 human participants in a live setting, but it does not refer to a publicly available or open dataset that would require specific access information. |
| Dataset Splits | No | The paper describes an interactive robotic experiment setup with human participants and does not mention train/validation/test dataset splits, as it does not rely on a pre-collected dataset for evaluation. |
| Hardware Specification | Yes | We have used a Softbank Robotics Pepper robot. Two Intel Real Sense D435i RGB+D cameras are used to provide high-resolution depth imagery. |
| Software Dependencies | No | The paper mentions several software components like Open Pose, dlib CNN, April Tag, and Dan Speech, but it does not provide specific version numbers for any of them. |
| Experiment Setup | No | The paper describes the physical setup of the robot, cameras, and objects, and the logic for event detection and epistemic model updates. However, it does not provide specific hyperparameter values, model initialization details, or training schedules for any of its components or for the overall system. |