{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,5]],"date-time":"2026-03-05T14:29:50Z","timestamp":1772720990572,"version":"3.50.1"},"reference-count":33,"publisher":"MDPI AG","issue":"8","license":[{"start":{"date-parts":[[2025,8,9]],"date-time":"2025-08-09T00:00:00Z","timestamp":1754697600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001691","name":"Japan Society for the Promotion of Science (JSPS) KAKENHI, Transformative Research Area (A): unified theory of prediction and action","doi-asserted-by":"publisher","award":["24H02175"],"award-info":[{"award-number":["24H02175"]}],"id":[{"id":"10.13039\/501100001691","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001691","name":"Japan Society for the Promotion of Science (JSPS) KAKENHI, Transformative Research Area (A): unified theory of prediction and action","doi-asserted-by":"publisher","award":["JPMJPR22C9"],"award-info":[{"award-number":["JPMJPR22C9"]}],"id":[{"id":"10.13039\/501100001691","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002241","name":"JST PRESTO","doi-asserted-by":"publisher","award":["24H02175"],"award-info":[{"award-number":["24H02175"]}],"id":[{"id":"10.13039\/501100002241","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002241","name":"JST PRESTO","doi-asserted-by":"publisher","award":["JPMJPR22C9"],"award-info":[{"award-number":["JPMJPR22C9"]}],"id":[{"id":"10.13039\/501100002241","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Entropy"],"abstract":"<jats:p>Active inference offers a unified framework in which agents can exhibit both goal-directed and epistemic behaviors. However, implementing policy search in high-dimensional continuous action spaces presents challenges in terms of scalability and stability. Our previously proposed model, T-GLean, addressed this issue by enabling efficient goal-directed planning through low-dimensional latent space search, further reduced by conditioning on prior habituated behavior. However, the lack of an epistemic term in minimizing expected free energy limited the agent\u2019s ability to engage in information-seeking behavior that can be critical for attaining preferred outcomes. In this study, we present EFE-GLean, an extended version of T-GLean that overcomes this limitation by integrating epistemic value into the planning process. EFE-GLean generates goal-directed policies by inferring low-dimensional future posterior trajectories while maximizing expected information gain. Simulation experiments using an extended T-maze task\u2014implemented in both discrete and continuous domains\u2014demonstrate that the agent can successfully achieve its goals by exploiting hidden environmental information. Furthermore, we show that the agent is capable of adapting to abrupt environmental changes by dynamically revising plans through simultaneous minimization of past variational free energy and future expected free energy. Finally, analytical evaluations detail the underlying mechanisms and computational properties of the model.<\/jats:p>","DOI":"10.3390\/e27080846","type":"journal-article","created":{"date-parts":[[2025,8,11]],"date-time":"2025-08-11T11:02:09Z","timestamp":1754910129000},"page":"846","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Active Inference with Dynamic Planning and Information Gain in Continuous Space by Inferring Low-Dimensional Latent States"],"prefix":"10.3390","volume":"27","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-1388-2906","authenticated-orcid":false,"given":"Takazumi","family":"Matsumoto","sequence":"first","affiliation":[{"name":"Cognitive Neurorobotics Research Unit, Okinawa Institute of Science and Technology, Okinawa 904-0495, Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-8906-0609","authenticated-orcid":false,"given":"Kentaro","family":"Fujii","sequence":"additional","affiliation":[{"name":"Department of Electronics and Electrical Engineering, Keio University, Kanagawa 223-8522, Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0790-1915","authenticated-orcid":false,"given":"Shingo","family":"Murata","sequence":"additional","affiliation":[{"name":"Department of Electronics and Electrical Engineering, Keio University, Kanagawa 223-8522, Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9131-9206","authenticated-orcid":false,"given":"Jun","family":"Tani","sequence":"additional","affiliation":[{"name":"Cognitive Neurorobotics Research Unit, Okinawa Institute of Science and Technology, Okinawa 904-0495, Japan"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,8,9]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"815","DOI":"10.1098\/rstb.2005.1622","article-title":"A theory of cortical responses","volume":"360","author":"Friston","year":"2005","journal-title":"Philos. Trans. R. Soc. B Biol. Sci."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1080\/17588928.2015.1020053","article-title":"Active inference and epistemic value","volume":"6","author":"Friston","year":"2015","journal-title":"Cogn. Neurosci."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"495","DOI":"10.1007\/s00422-019-00805-w","article-title":"Generalised free energy and active inference","volume":"113","author":"Parr","year":"2019","journal-title":"Biol. Cybern."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Matsumoto, T., Ohata, W., Benureau, F.C.Y., and Tani, J. (2022). Goal-Directed Planning and Goal Understanding by Extended Active Inference: Evaluation through Simulated and Physical Robot Experiments. Entropy, 24.","DOI":"10.3390\/e24040469"},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Matsumoto, T., Ohata, W., and Tani, J. (2023). Incremental Learning of Goal-Directed Actions in a Dynamic Environment by a Robot Using Active Inference. Entropy, 25.","DOI":"10.3390\/e25111506"},{"key":"ref_6","unstructured":"Fountas, Z., Sajid, N., Mediano, P.A., and Friston, K.J. (2020, January 6\u201312). Deep active inference agents using Monte-Carlo methods. Proceedings of the 34th International Conference on Neural Information Processing Systems, Vancouver, BC, Canada."},{"key":"ref_7","unstructured":"Fedosov, N., and Voskoboynikov, A. (2023, January 28\u201330). Deep Active Inference Agent with Continuous Action Space. Proceedings of the 2023 IEEE Ural-Siberian Conference on Computational Technologies in Cognitive Science, Genomics and Biomedicine (CSGB), Novosibirsk, Russia."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Cogliati Dezza, I., Schulz, E., and Wu, C.M. (2022). Active Inference, Bayesian Optimal Design, and Expected Utility. The Drive for Knowledge: The Science of Human Information Seeking, Cambridge University Press.","DOI":"10.1017\/9781009026949"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Champion, T., Grze\u015b, M., Bonheme, L., and Bowman, H. (2023). Deconstructing deep active inference. arXiv.","DOI":"10.2139\/ssrn.4398264"},{"key":"ref_10","first-page":"70","article-title":"A free energy principle for the brain","volume":"100","author":"Friston","year":"2006","journal-title":"J. Physiol."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"447","DOI":"10.1162\/neco_a_01354","article-title":"Whence the Expected Free Energy?","volume":"33","author":"Millidge","year":"2021","journal-title":"Neural Comput."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"137","DOI":"10.1007\/s00422-011-0424-z","article-title":"Action understanding and active inference","volume":"104","author":"Friston","year":"2011","journal-title":"Biol. Cybern."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"547","DOI":"10.1007\/s00422-018-0785-7","article-title":"Deep active inference","volume":"112","year":"2018","journal-title":"Biol. Cybern."},{"key":"ref_14","unstructured":"Yeganeh, Y.T., Jafari, M., and Matta, A. (2025). Deep Active Inference Agents for Delayed and Long-Horizon Environments. arXiv."},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Igari, K., Fujii, K., Haddon-Hill, G.W., and Murata, S. (2024, January 9\u201311). Selection of Exploratory or Goal-Directed Behavior by a Physical Robot Implementing Deep Active Inference. Proceedings of the 5th International Workshop on Active Inference, IWAI 2024, Oxford, UK.","DOI":"10.1007\/978-3-031-77138-5_11"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Brown, H., and Friston, K. (2012). Free-energy and illusions: The Cornsweet effect. Front. Psychol., 3.","DOI":"10.3389\/fpsyg.2012.00043"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1162\/NECO_a_00912","article-title":"Active Inference: A Process Theory","volume":"29","author":"Friston","year":"2017","journal-title":"Neural Comput."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"2973","DOI":"10.1109\/LRA.2020.2974451","article-title":"A Novel Adaptive Controller for Robot Manipulators Based on Active Inference","volume":"5","author":"Pezzato","year":"2020","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_19","unstructured":"Hafner, D., Lillicrap, T., Fischer, I., Villegas, R., Ha, D., Lee, H., and Davidson, J. (2019, January 10\u201315). Learning latent dynamics for planning from pixels. Proceedings of the International Conference on Machine Learning, PMLR, Long Beach, CA, USA."},{"key":"ref_20","unstructured":"Hafner, D., Lillicrap, T.P., Norouzi, M., and Ba, J. (2021, January 3\u20137). Mastering Atari with Discrete World Models. Proceedings of the International Conference on Learning Representations, Virtual."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"2025","DOI":"10.1162\/neco_a_01228","article-title":"A novel predictive-coding-inspired variational RNN model for online prediction and recognition","volume":"31","author":"Ahmadi","year":"2019","journal-title":"Neural Comput."},{"key":"ref_22","unstructured":"Kingma, D.P., and Welling, M. (2014, January 14\u201316). Auto-Encoding Variational Bayes. Proceedings of the 2nd International Conference on Learning Representations, ICLR 2014, Banff, AB, Canada."},{"key":"ref_23","unstructured":"Rumelhart, D.E., Hinton, G.E., and Williams, R.J. (1986). Learning internal representations by error propagation. Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Vol. 1: Foundations, MIT Press."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Matsumoto, T., and Tani, J. (2020). Goal-Directed Planning for Habituated Agents by Active Inference Using a Variational Recurrent Neural Network. Entropy, 22.","DOI":"10.3390\/e22050564"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Friston, K., Schwartenbeck, P., FitzGerald, T., Moutoussis, M., Behrens, T., and Dolan, R.J. (2014). The anatomy of choice: Dopamine and decision-making. Philos. Trans. R. Soc. B Biol. Sci., 369.","DOI":"10.1098\/rstb.2013.0481"},{"key":"ref_26","unstructured":"Oliver, G., Lanillos, P., and Cheng, G. (2019). Active inference body perception and action for humanoid robots. arXiv."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"462","DOI":"10.1109\/TCDS.2021.3049907","article-title":"An Empirical Study of Active Inference on a Humanoid Robot","volume":"14","author":"Oliver","year":"2021","journal-title":"IEEE Trans. Cogn. Dev. Syst."},{"key":"ref_28","doi-asserted-by":"crossref","unstructured":"Meo, C., and Lanillos, P. (October, January 27). Multimodal VAE Active Inference Controller. Proceedings of the 2021 IEEE\/RSJ International Conference on Intelligent Robots and Systems (IROS), Prague, Czech Republic.","DOI":"10.1109\/IROS51168.2021.9636394"},{"key":"ref_29","unstructured":"Bellec, G., Scherr, F., Hajek, E., Salaj, D., Legenstein, R., and Maass, W. (2019). Biologically inspired alternatives to backpropagation through time for learning in recurrent neural nets. arXiv."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"659","DOI":"10.1007\/978-3-030-61616-8_53","article-title":"Learning Precise Spike Timings with Eligibility Traces","volume":"Volume 12397","author":"Masulli","year":"2020","journal-title":"Artificial Neural Networks and Machine Learning\u2014ICANN 2020"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"1329","DOI":"10.1162\/neco_a_01497","article-title":"Predictive Coding Approximates Backprop Along Arbitrary Computation Graphs","volume":"34","author":"Millidge","year":"2022","journal-title":"Neural Comput."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"49","DOI":"10.1016\/j.robot.2012.05.008","article-title":"Active learning of inverse models with intrinsically motivated goal exploration in robots","volume":"61","author":"Baranes","year":"2013","journal-title":"Robot. Auton. Syst."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Parr, T., and Friston, K.J. (2017). Uncertainty, epistemics and active inference. J. R. Soc. Interface, 14.","DOI":"10.1098\/rsif.2017.0376"}],"container-title":["Entropy"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1099-4300\/27\/8\/846\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,9]],"date-time":"2025-10-09T18:27:10Z","timestamp":1760034430000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1099-4300\/27\/8\/846"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,9]]},"references-count":33,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2025,8]]}},"alternative-id":["e27080846"],"URL":"https:\/\/doi.org\/10.3390\/e27080846","relation":{},"ISSN":["1099-4300"],"issn-type":[{"value":"1099-4300","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,9]]}}}