Kolby Nottingham
Description: Embodied Decision Making using Language Guided World Modelling
Reinforcement learning (RL) agents typically learn tabula rasa, without prior knowledge of the world. However, if initialized with knowledge of high-level subgoals and transitions between subgoals, RL agents could utilize this Abstract World Model (AWM) for planning and exploration. We propose using few-shot large language models (LLMs) to hypothesize an AWM, that will be verified through world experience, to improve sample efficiency of RL agents. Our DECKARD agent applies LLM-guided exploration to item cr
Paper» Code» Video»
We present DECKARD ( DEC ision-making for K nowledgable A utonomous R einforcement-leanring D reamers), an agent that hypothesizes an Abstract World Model (AWM) over subgoals by few-shot prompting an LLM, then exploits the AWM for exploration and verifies the AWM with grounded experience.
Kolby Nottingham