Nov 8, 2022 · In this work, we study the ability of RL agents to follow such instructions. We conduct experiments that show that the performance of state-of- ...
In this paper we study instruction following in text-based games and propose an approach that advances the previous state of the art. To this end, we employ the ...
Oct 31, 2022 · The performance of a reinforcement learning system for playing text-based games is improved by converting in-game instructions into Linear Temporal Logic (LTL).
Reinforcement Learning for Text-Based Games. - Partial observability. - memory, exploration, navigation. - Language understanding. - Instruction following. Page ...
Nov 8, 2022 · In this work, we study the ability of RL agents to follow such instructions. We conduct experiments that show that the performance of state-of- ...
Apr 3, 2024 · In this work, we study the ability of RL agents to follow such instructions. We conduct experiments that show that the performance of state-of-the-art text- ...
Translate natural language instructions to Linear Temporal Logic (LTL). ○ express complex instructions compactly. ○ exploit LTL's compositional syntax and ...
This is the code repository for the paper Learning to Follow Instruction in Text-Based Games, which will appear in NeurIPS 2022. Installing Dependencies. We ...
We build a simple translator that reads game observations and constructs these LTL instructions directly, but only once. Repeated observations will not result ...
Learning to Follow Instructions in Text-Based Games. Tuli, M., Li, A ... . Learning to Follow Instructions in Text-Based Games [pdf] Paper bibtex 73 downloads