高技术通讯2024，Vol.34Issue(6)：555-566,12.DOI:10.3772/j.issn.1002-0470.2024.06.001
基于语言类任务的概念化强化学习框架
Conceptual reinforcement learning for language-assisted tasks
摘要
Abstract
Language-assisted tasks are proposed to facilitate the generalization ability of reinforcement learning policy.The key question is to learn the general representation across different scenarios.Existing studies often implicitly learn the joint representation,which may include spurious correlation information and consequently compromise pol-icy's generalization performance and training efficiency.To address this issue,a conceptual reinforcement learning framework(CRL)is proposed,which exploits the motivation of human cognition that extracts similarits from nu-merous instances to generate conceptual abstraction,and incorporates a multi-level attention encoder and restricted loss functions to learn compact and invariant conceptual representation for the policy.Evaluated in challenging lan-guage-assisted tasks,the results demonstrate that CRL significantly improves the policy's training efficiency(up to 70%)and generalization ability(up to30%).Additionally,the conceptual representation also shows better inter-pretability than other representations.关键词
深度强化学习(DRL)/语言类强化学习任务/文本游戏/表示学习/互信息优化Key words
deep reinforcement learning(DRL)/language-assisted reinforcement learning task/text game/representation learning/mutual information引用本文复制引用
彭少辉,胡杏,支天..基于语言类任务的概念化强化学习框架[J].高技术通讯,2024,34(6):555-566,12.基金项目
国家自然科学基金(62002338,U20A20227,U22A2028)和中国科学院稳定支持基础研究领域青年团队计划(YSBR-029)资助项目. （62002338,U20A20227,U22A2028）
