{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T15:34:32Z","timestamp":1759332872140,"version":"build-2065373602"},"reference-count":41,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"10","license":[{"start":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T00:00:00Z","timestamp":1759276800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T00:00:00Z","timestamp":1759276800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,10,1]],"date-time":"2025-10-01T00:00:00Z","timestamp":1759276800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Inform. Theory"],"published-print":{"date-parts":[[2025,10]]},"DOI":"10.1109\/tit.2025.3598296","type":"journal-article","created":{"date-parts":[[2025,8,13]],"date-time":"2025-08-13T17:34:10Z","timestamp":1755106450000},"page":"7942-7958","source":"Crossref","is-referenced-by-count":0,"title":["Sample-Efficient Reinforcement Learning From Human Feedback via Information-Directed Sampling"],"prefix":"10.1109","volume":"71","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-7684-9882","authenticated-orcid":false,"given":"Han","family":"Qi","sequence":"first","affiliation":[{"name":"School of Software Engineering, Xi’an Jiaotong University, Xi’an, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-1772-4172","authenticated-orcid":false,"given":"Haochen","family":"Yang","sequence":"additional","affiliation":[{"name":"School of Mathematics, Peking University, Beijing, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6114-8453","authenticated-orcid":false,"given":"Qiaosheng","family":"Zhang","sequence":"additional","affiliation":[{"name":"Shanghai Artificial Intelligence Laboratory, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhuoran","family":"Yang","sequence":"additional","affiliation":[{"name":"Department of Statistics and Data Science, Yale University, New Haven, CT, USA"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","first-page":"27730","article-title":"Training language models to follow instructions with human feedback","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Ouyang"},{"key":"ref2","first-page":"43037","article-title":"Principled reinforcement learning with human feedback from pairwise or K-wise comparisons","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhu"},{"key":"ref3","first-page":"3773","article-title":"Human-in-the-loop: Provably efficient preference-based reinforcement learning with general function approximation","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Chen"},{"key":"ref4","article-title":"Exploratory preference optimization: Harnessing implicit Q-approximation for sample-efficient RLHF","author":"Xie","year":"2024","journal-title":"arXiv:2405.21046"},{"key":"ref5","article-title":"Near-optimal optimistic reinforcement learning using empirical Bernstein inequalities","author":"Tossou","year":"2019","journal-title":"arXiv:1905.12425"},{"key":"ref6","article-title":"Online iterative reinforcement learning from human feedback with general preference model","author":"Ye","year":"2024","journal-title":"arXiv:2402.07314"},{"key":"ref7","first-page":"23557","article-title":"Improved Bayesian regret bounds for Thompson sampling in reinforcement learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Moradipari"},{"key":"ref8","article-title":"Making RL with preference-based feedback efficient via randomization","author":"Wu","year":"2023","journal-title":"arXiv:2310.14554"},{"key":"ref9","first-page":"28575","article-title":"Regret bounds for information-directed reinforcement learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Hao"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1287\/opre.2017.1663"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1016\/j.artint.2025.104392"},{"key":"ref12","first-page":"16738","article-title":"Information directed sampling for sparse linear bandits","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Hao"},{"key":"ref13","first-page":"8446","article-title":"Contextual information-directed sampling","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Hao"},{"key":"ref14","article-title":"Proximal policy optimization algorithms","author":"Schulman","year":"2017","journal-title":"arXiv:1707.06347"},{"key":"ref15","article-title":"GPT-4 technical report","volume-title":"arXiv:2303.08774","author":"Achiam","year":"2023"},{"key":"ref16","article-title":"Llama 2: Open foundation and fine-tuned chat models","author":"Touvron","year":"2023","journal-title":"arXiv:2307.09288"},{"key":"ref17","article-title":"Fine-tuning language models from human preferences","author":"Ziegler","year":"2019","journal-title":"arXiv:1909.08593"},{"key":"ref18","first-page":"53728","article-title":"Direct preference optimization: Your language model is secretly a reward model","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Rafailov"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1007\/s10994-012-5313-8"},{"key":"ref20","first-page":"30050","article-title":"Optimal algorithms for stochastic contextual preference bandits","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"34","author":"Saha"},{"key":"ref21","article-title":"Reinforcement learning from human feedback with active queries","author":"Ji","year":"2024","journal-title":"arXiv:2402.09401"},{"key":"ref22","first-page":"11261","article-title":"Contextual bandits and imitation learning with preference-based active queries","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","volume":"36","author":"Sekhari"},{"key":"ref23","article-title":"Feel-good Thompson sampling for contextual dueling bandits","author":"Li","year":"2024","journal-title":"arXiv:2404.06013"},{"article-title":"Online preference alignment for language models via count-based exploration","volume-title":"Proc. Int. Conf. Learn. Represent. (ICLR)","author":"Bai","key":"ref24"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1007\/s10994-014-5458-8"},{"key":"ref26","first-page":"18784","article-title":"Preference-based reinforcement learning with finite-time guarantees","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Xu"},{"key":"ref27","article-title":"Dueling RL: Reinforcement learning with trajectory preferences","author":"Pacchiano","year":"2021","journal-title":"arXiv:2111.04850"},{"article-title":"Adversarial imitation learning with preferences","volume-title":"Proc. The 11th Int. Conf. Learn. Represent.","author":"Taranovic","key":"ref28"},{"issue":"1","key":"ref29","first-page":"2442","article-title":"An information-theoretic analysis of Thompson sampling","volume":"17","author":"Russo","year":"2016","journal-title":"J. Mach. Learn. Res."},{"key":"ref30","first-page":"4161","article-title":"An information-theoretic analysis for Thompson sampling with many actions","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Dong"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/tit.2022.3213630"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.11751"},{"key":"ref33","first-page":"2777","article-title":"Asymptotically optimal information-directed sampling","volume-title":"Proc. Conf. Learn. Theory","author":"Kirschner"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.2307\/2334029"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/ICRA46639.2022.9811590"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1016\/j.robot.2023.104431"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1111\/tops.12525"},{"key":"ref38","first-page":"6263","article-title":"Dueling RL: Reinforcement learning with trajectory preferences","volume-title":"Proc. Int. Conf. Artif. Intell. Statist.","author":"Saha"},{"key":"ref39","first-page":"3003","article-title":"(More) efficient reinforcement learning via posterior sampling","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Osband"},{"key":"ref40","article-title":"The statistical complexity of interactive decision making","author":"Foster","year":"2021","journal-title":"arXiv:2112.13487"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1017\/9781139029834"}],"container-title":["IEEE Transactions on Information Theory"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/18\/11175289\/11123904.pdf?arnumber=11123904","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,30]],"date-time":"2025-09-30T16:09:28Z","timestamp":1759248568000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11123904\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10]]},"references-count":41,"journal-issue":{"issue":"10"},"URL":"https:\/\/doi.org\/10.1109\/tit.2025.3598296","relation":{},"ISSN":["0018-9448","1557-9654"],"issn-type":[{"type":"print","value":"0018-9448"},{"type":"electronic","value":"1557-9654"}],"subject":[],"published":{"date-parts":[[2025,10]]}}}