Skill Machines: Temporal Logic Skill Composition in Reinforcement Learning скачать в хорошем качестве

Skill Machines: Temporal Logic Skill Composition in Reinforcement Learning

KmerAI Talks#04 with Geraud Nangue Tasse About the Guest: Geraud is an IBM PhD fellow, PhD candidate, and lecturer in Computer Science at the University of the Witwatersrand. Geraud has also done a summer research internship at IBM Thomas J. Watson Research Center in New York in 2022 under the course of his PhD. More information about the speaker can be found here: https://geraudnt.github.io/. Topic: Skill Machines: Temporal Logic Skill Composition in Reinforcement Learning Geraud will discuss the challenges associated with the capability of agents to solve a diverse range of problems specified via language within the same environment. He will introduce their methodology for addressing this challenge, showing how the method allows the agent to map complex task specifications to near-optimal behaviors without additional training. He will present empirical results on various datasets including a tabular setting, a high-dimensional video game setting, and a continuous control environment. Checkout the paper: https://arxiv.org/abs/2205.12532