Papers by Michael Tarr
Open-Ended Instructable Embodied Agents with Memory-Augmented Large Language Models (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Pre-trained and frozen LLMs can effectively map simple scene re-arrangement instructions to programs over a robot’s visuomotor functions, but fixed prompts fall short. |
| Approach: | They propose an embodied agent equipped with an external memory of language-program pairs that parses free-form human-robot dialogue into action programs through retrieval-augmented LLM prompting. |
| Outcome: | The proposed agent parses human-robot dialogue into action programs using retrieval-augmented LLM prompting. |