{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,15]],"date-time":"2026-07-15T17:13:49Z","timestamp":1784135629754,"version":"3.55.0"},"reference-count":76,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"11","license":[{"start":{"date-parts":[[2025,11,1]],"date-time":"2025-11-01T00:00:00Z","timestamp":1761955200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,11,1]],"date-time":"2025-11-01T00:00:00Z","timestamp":1761955200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,11,1]],"date-time":"2025-11-01T00:00:00Z","timestamp":1761955200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Science Foundation of China","doi-asserted-by":"publisher","award":["62276126"],"award-info":[{"award-number":["62276126"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Science Foundation of China","doi-asserted-by":"publisher","award":["62250069"],"award-info":[{"award-number":["62250069"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Neural Netw. Learning Syst."],"published-print":{"date-parts":[[2025,11]]},"DOI":"10.1109\/tnnls.2025.3590731","type":"journal-article","created":{"date-parts":[[2025,8,14]],"date-time":"2025-08-14T18:48:20Z","timestamp":1755197300000},"page":"19681-19692","source":"Crossref","is-referenced-by-count":3,"title":["Improving Sample Efficiency of Reinforcement Learning With Background Knowledge From Large Language Models"],"prefix":"10.1109","volume":"36","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-7251-4372","authenticated-orcid":false,"given":"Fuxiang","family":"Zhang","sequence":"first","affiliation":[{"name":"National Key Laboratory for Novel Software Technology and School of Artificial Intelligence, Nanjing University, Nanjing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Junyou","family":"Li","sequence":"additional","affiliation":[{"name":"Tencent, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-9908-5303","authenticated-orcid":false,"given":"Yi-Chen","family":"Li","sequence":"additional","affiliation":[{"name":"National Key Laboratory for Novel Software Technology and School of Artificial Intelligence, Nanjing University, Nanjing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9238-4747","authenticated-orcid":false,"given":"Zongzhang","family":"Zhang","sequence":"additional","affiliation":[{"name":"National Key Laboratory for Novel Software Technology and School of Artificial Intelligence, Nanjing University, Nanjing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1052-5447","authenticated-orcid":false,"given":"Yang","family":"Yu","sequence":"additional","affiliation":[{"name":"National Key Laboratory for Novel Software Technology and School of Artificial Intelligence, Nanjing University, Nanjing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1754-1837","authenticated-orcid":false,"given":"Deheng","family":"Ye","sequence":"additional","affiliation":[{"name":"Tencent, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","volume-title":"Reinforcement Learning: An Introduction","author":"Sutton","year":"2018"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1038\/nature14236"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.13140\/RG.2.2.18893.74727"},{"key":"ref4","first-page":"621","article-title":"Towards playing full MOBA games with deep reinforcement learning","volume-title":"Proc. Conf. Neural Inf. Process. Syst.","volume":"33","author":"Ye"},{"key":"ref5","article-title":"Solving Rubik\u2019s cube with a robot hand","author":"Akkaya","year":"2019","journal-title":"arXiv:1910.07113"},{"key":"ref6","first-page":"27730","article-title":"Training language models to follow instructions with human feedback","volume-title":"Proc. Adv. Neural Inf. Process. Syst. (NIPS)","author":"Ouyang"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2018\/820"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2022.03.003"},{"key":"ref9","article-title":"A survey on intrinsic motivation in reinforcement learning","author":"Aubret","year":"2019","journal-title":"arXiv:1908.06976"},{"key":"ref10","first-page":"1471","article-title":"Unifying count-based exploration and intrinsic motivation","volume-title":"Proc. Int. Conf. Adv. Neural Inf. Process. Syst.","author":"Bellemare"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2017.70"},{"key":"ref12","article-title":"Exploration by random network distillation","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Burda"},{"key":"ref13","first-page":"25217","article-title":"NovelD: A simple yet effective exploration criterion","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Zhang"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2019\/331"},{"key":"ref15","article-title":"Knowledge-guided exploration in deep reinforcement learning","author":"Mazumder","year":"2022","journal-title":"arXiv:2210.15670"},{"key":"ref16","article-title":"Reinforcement learning with foundation priors: Let the embodied agent efficiently learn on its own","author":"Ye","year":"2023","journal-title":"arXiv:2310.02635"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-024-40231-1"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-023-2689-5"},{"key":"ref19","article-title":"A survey of large language models","author":"Xin Zhao","year":"2023","journal-title":"arXiv:2303.18223"},{"key":"ref20","article-title":"Sparks of artificial general intelligence: Early experiments with GPT-4","author":"Bubeck","year":"2023","journal-title":"arXiv:2303.12712"},{"key":"ref21","article-title":"GPT-4 technical report","volume-title":"arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"ref22","article-title":"Towards a unified agent with foundation models","volume-title":"Proc. Workshop Reincarnating Reinforcement Learn. ICLR","author":"Palo"},{"key":"ref23","first-page":"8657","article-title":"Guiding pretraining in reinforcement learning with large language models","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Du"},{"key":"ref24","first-page":"31967","article-title":"Large language models as commonsense knowledge for large-scale task planning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Zhao"},{"key":"ref25","article-title":"Eureka: Human-level reward design via coding large language models","author":"Jason Ma","year":"2023","journal-title":"arXiv:2310.12931"},{"key":"ref26","first-page":"374","article-title":"Language to rewards for robotic skill synthesis","volume-title":"Proc. Conf. Robot Learn.","author":"Yu"},{"key":"ref27","article-title":"Text2Reward: Reward shaping with language models for reinforcement learning","author":"Xie","year":"2023","journal-title":"arXiv:2309.11489"},{"key":"ref28","first-page":"278","article-title":"Policy invariance under reward transformations: Theory and application to reward shaping","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Ng"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-023-3150-5"},{"key":"ref30","first-page":"7754","article-title":"Generalized hindsight for reinforcement learning","volume-title":"Proc. NIPS","author":"Li"},{"key":"ref31","article-title":"A narration-based reward shaping approach using grounded natural language commands","author":"Waytowich","year":"2019","journal-title":"arXiv:1911.00497"},{"key":"ref32","first-page":"33947","article-title":"Improving intrinsic exploration with language abstractions","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Mu"},{"key":"ref33","first-page":"25377","article-title":"Semantic exploration from language abstractions and pretrained representations","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Tam"},{"key":"ref34","first-page":"29529","article-title":"ELLA: Exploration through learned language abstraction","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Mirchandani"},{"key":"ref35","first-page":"12478","article-title":"EAGER: Asking and answering questions for automatic reward shaping in language-guided RL","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Carta"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1145\/3580305.3599393"},{"key":"ref37","first-page":"1877","article-title":"Language models are few-shot learners","volume-title":"Proc. NIPS","author":"Brown"},{"key":"ref38","first-page":"24824","article-title":"Chain-of-thought prompting elicits reasoning in large language models","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Lee"},{"key":"ref39","article-title":"Voyager: An open-ended embodied agent with large language models","author":"Wang","year":"2023","journal-title":"arXiv:2305.16291"},{"key":"ref40","article-title":"SPRING: Studying the paper and reasoning to play games","author":"Wu","year":"2023","journal-title":"arXiv:2305.15486"},{"key":"ref41","first-page":"287","article-title":"Do as I can, not as I say: Grounding language in robotic affordances","volume-title":"Proc. Conf. Robot Learn.","author":"Ahn"},{"key":"ref42","first-page":"31199","article-title":"Pre-trained language models for interactive decision-making","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Li"},{"key":"ref43","first-page":"3676","article-title":"Grounding large language models in interactive environments with online reinforcement learning","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Carta"},{"key":"ref44","article-title":"Enabling intelligent interactions between an agent and an LLM: A reinforcement learning approach","author":"Hu","year":"2023","journal-title":"arXiv:2306.03604"},{"key":"ref45","article-title":"LaGR-SEQ: Language-guided reinforcement learning with sample-efficient querying","author":"George Karimpanal","year":"2023","journal-title":"arXiv:2308.13542"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2019\/880"},{"key":"ref47","article-title":"Reward design with language models","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Kwon"},{"key":"ref48","article-title":"Flexible attention-based multi-policy fusion for efficient deep reinforcement learning","author":"Chiu","year":"2022","journal-title":"arXiv:2210.03729"},{"key":"ref49","article-title":"Knowledge graph injection for reinforcement learning","volume-title":"Proc. Workshop Deep Learn. Knowl. Graphs 21st Int. Semantic Web Conf.","author":"Wardenga"},{"key":"ref50","article-title":"Efficient reinforcement learning with large language model priors","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Xue"},{"key":"ref51","article-title":"Sample efficient preference alignment in LLMs via active exploration","author":"Mehta","year":"2023","journal-title":"arXiv:2312.00267"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i10.28953"},{"key":"ref53","first-page":"463","article-title":"Learning to drive a bicycle using reinforcement learning and shaping","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Randl\u00f8v"},{"key":"ref54","article-title":"D4RL: Datasets for deep data-driven reinforcement learning","author":"Fu","year":"2020","journal-title":"arXiv:2004.07219"},{"key":"ref55","article-title":"Proximal policy optimization algorithms","author":"Schulman","year":"2017","journal-title":"arXiv:1707.06347"},{"key":"ref56","first-page":"1","article-title":"CleanRL: High-quality single-file implementations of deep reinforcement learning algorithms","author":"Huang","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref57","article-title":"Evaluating large language models trained on code","author":"Chen","year":"2021","journal-title":"arXiv:2107.03374"},{"key":"ref58","article-title":"CodeGen: An open large language model for code with multi-turn program synthesis","author":"Nijkamp","year":"2022","journal-title":"arXiv:2203.13474"},{"key":"ref59","article-title":"If LLM is the wizard, then code is the wand: A survey on how code empowers large language models to serve as intelligent agents","author":"Yang","year":"2024","journal-title":"arXiv:2401.00812"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.nlp4convai-1.5"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.acl-long.870"},{"key":"ref62","article-title":"Constitutional AI: Harmlessness from AI feedback","author":"Bai","year":"2022","journal-title":"arXiv:2212.08073"},{"key":"ref63","article-title":"RLAIF vs. RLHF: Scaling reinforcement learning from human feedback with AI feedback","author":"Lee","year":"2023","journal-title":"arXiv:2309.00267"},{"key":"ref64","article-title":"Motif: Intrinsic motivation from artificial intelligence feedback","author":"Klissarov","year":"2023","journal-title":"arXiv:2310.00166"},{"key":"ref65","first-page":"1133","article-title":"A Bayesian approach for policy learning from trajectory preference queries","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"25","author":"Wilson"},{"key":"ref66","first-page":"4299","article-title":"Deep reinforcement learning from human preferences","volume-title":"Proc. Int. Conf. Adv. Neural Inf. Process. Syst.","volume":"30","author":"Christiano"},{"key":"ref67","doi-asserted-by":"publisher","DOI":"10.2307\/2334029"},{"key":"ref68","article-title":"Preference transformer: Modeling human preferences using transformers for RL","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Kim"},{"key":"ref69","article-title":"Semantically aligned task decomposition in multi-agent reinforcement learning","author":"Li","year":"2023","journal-title":"arXiv:2305.10865"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-1410"},{"key":"ref71","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.04805"},{"key":"ref72","article-title":"Minigrid & miniworld: Modular & customizable reinforcement learning environments for goal-oriented tasks","author":"Chevalier-Boisvert","year":"2023","journal-title":"arXiv:2306.13831"},{"key":"ref73","article-title":"Benchmarking the spectrum of agent capabilities","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Hafner"},{"key":"ref74","article-title":"SmartPlay: A benchmark for LLMs as intelligent agents","author":"Wu","year":"2023","journal-title":"arXiv:2310.01557"},{"key":"ref75","article-title":"BabyAI: A platform to study the sample efficiency of grounded language learning","volume-title":"Proc. Int. Conf. Learn. Represent.","author":"Chevalier-Boisvert"},{"key":"ref76","article-title":"Language model self-improvement by reinforcement learning contemplation","author":"Pang","year":"2023","journal-title":"arXiv:2305.14483"}],"container-title":["IEEE Transactions on Neural Networks and Learning Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/5962385\/11220834\/11125495.pdf?arnumber=11125495","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,30]],"date-time":"2025-10-30T18:03:26Z","timestamp":1761847406000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11125495\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11]]},"references-count":76,"journal-issue":{"issue":"11"},"URL":"https:\/\/doi.org\/10.1109\/tnnls.2025.3590731","relation":{},"ISSN":["2162-237X","2162-2388"],"issn-type":[{"value":"2162-237X","type":"print"},{"value":"2162-2388","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11]]}}}