The DOI matches its Crossref record; 73 of the 78 references it cites were found in a registry or an index.
Deep active inference
Open accessAbstract
This work combines the free energy principle and the ensuing active inference dynamics with recent advances in variational inference in deep generative models, and evolution strategies to introduce the "deep active inference" agent. This agent minimises a variational free energy bound on the average surprise of its sensations, which is motivated by a homeostatic argument. It does so by optimising the parameters of a generative latent variable model of its sensory inputs, together with a variational density approximating the posterior distribution over the latent variables, given its observations, and by acting on its environment to actively sample input that is likely under this generative model. The internal dynamics of the agent are implemented using deep and recurrent neural networks, as used in machine learning, making the deep active inference agent a scalable and very flexible class of active inference agent. Using the mountain car problem, we show how goal-directed behaviour can be implemented by defining appropriate priors on the latent states in the agent's model. Furthermore, we show that the deep active inference agent can learn a generative model of the environment, which can be sampled from to understand the agent's beliefs about the environment and its interaction therewith.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“First, there is no expected free energy, instead the agent is trained to minimize the cumulated variational free energy over time.”
From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Ueltzhöffer (2018)
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Third, the DAI F A requires the modeller to encode the prior preferences within the distributions predicted by the encoder and transition network.This can limit the applicability of the approach.”
From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Ueltzhöffer (2018)
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Second, this approach unrolls the partially observable Markov decision process over time.In other words, the code builds a huge computational graph containing the encoder, decoder, transition and policy networks for each action-perception cycle.Therefore, the approach is computationally intensive and can become intractable for a large time horizon.”
From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Ueltzhöffer (2018)
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Although these packages are designed explicitly for reinforcement learning, which involves computing a reward metric at every state, they can also be used for Active Inference as they describe general control environments [24,17].”
From Reactive Environments for Active Inference Agents with RxEnvironments.jl (2025) · cites this paper as [24]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“This effectively propels the agent towards the target location.”
From Deconstructing Deep Active Inference: A Contrarian Information Gatherer (2024) · cites this paper as Ueltzhöffer (2018)
4 more claims about this paper
Create a free account to see every claim other papers make about this one.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946 · cited 5×
“The chain rule gives: (2) Using these gradient estimates, we now optimise the expected free energy bound under the population density using ADAM as gradient based optimiser (Kingma and Ba, 2014).”
MethodInfluentialLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357
Unpaired image-to-image translation using cycle-consistent adversarial networks
Zhu, J.-Y., Park, T., Isola, P., & Efros, A. A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.244 · Cited by 23,602 · cited 1×
“2016) or style transfer or unsupervised image-to-image translation (Zhu et al. 2017; Liu et al. 2017).”
BackgroundMultilayer feedforward networks are universal approximators
Hornik, K., Stinchcombe, M., & White, H. (1989) · Neural Networks · doi:10.1016/0893-6080(89)90020-8 · Cited by 21,728
Reducing the dimensionality of data with neural networks
Hinton, G. E., & Salakhutdinov, R. R. (2006) · Science · doi:10.1126/science.1127647 · Cited by 20,928
Auto-Encoding Variational Bayes
Kingma, D. P., & Welling, M. (2013) · arXiv · doi:10.48550/arxiv.1312.6114 · Cited by 20,067 · cited 14×
“…Agent do for t = 1 , ..., T do Sample an action ˆ a t from p θ ( a t | ˆ s t − 1 ) Propagate the environment using where only one sample per data point is enough, since the gradients depend on entire (mini-)batches of datapoints (Kingma and Welling, 2014; Rezende et al, 2014; Chung et al, 2015).”
BackgroundMethodInfluential(2016) · Cited by 11,679
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., & WhiteH. (1989) · Neural Networks · doi:10.5555/70405.70408 · Cited by 9,081
The free-energy principle: a unified brain theory?
Friston, K. (2010) · Nature Reviews Neuroscience · doi:10.1038/nrn2787 · Cited by 8,120 · cited 2×
“…as well as anatomical and physiological features of the brain on one hand (Brown and Fris-ton, 2012; Friston, 2005; Friston et al, 2011; Friston and Kiebel, 2009; Schwartenbeck et al, 2015; Adams et al, 2013), and encompasses many formal theories about brain function on the other (Friston, 2010).”
BackgroundResult
Research works printed but not linked: 2
- Karras T, Aila T, Laine S, Lehtinen J (2018) Progressive growing of gans for improved quality, stability, and variation. ICLR
- Theano Development Team (2016) Theano: a Python framework for fast computation of mathematical expressions. arXiv:1605.02688 · arXiv
Cited by
The free energy principle made simpler but not too simple
Friston, K., Da Costa, L., Sajid, N., Heins, C., Ueltzhöffer, K., Pavliotis, G. A., & Parr, T. (2023) · Physics Reports · doi:10.1016/j.physrep.2023.07.001 · Cited by 163
“Recently, there has been an increasing focus on active inference in the setting of machine learning and artificial intelligence [61,147,150,182–185].”
Self-citationActive Inference: Demystified and Compared
Sajid, N., Ball, P. J., Parr, T., & Friston, K. J. (2021) · Neural Computation · doi:10.1162/neco_a_01357 · Cited by 252
“…2017; Friston, Parr, et al., 2017; Parr & Friston, 2017), continuous state-space models (Buckley, Kim, McGregor, & Seth, 2017; Parr & Friston, 2019a; Ueltzhöffer, 2018) or mixed models with both discrete and continuous states as described in Friston, Parr, et al. (2017) and Parr and Friston (2018).”
Background(2021) · UCL Discovery (University College London) · Cited by 141
Friston, K. J., Costa, L. D., Hafner, D., Hesp, C., & Parr, T. (2021) · Neural Computation · doi:10.1162/neco_a_01351 · Cited by 183
“For example, Ueltzhöffer (2018) parameterized both the generative model and approximate posterior with function approximators, using evolutionary schemes to minimize variational free energy when gradients were not available.”
MethodInfluentialDeep active inference as variational policy gradients
Millidge, B. (2020) · Journal of Mathematical Psychology · doi:10.1016/j.jmp.2020.102348 · Cited by 115
Metadata
- Authors
- Ueltzhöffer, K.
- Corresponding author
- Ueltzhöffer, Kai
- ORCID
- Ueltzhöffer, Kai
- Venue
- Biological Cybernetics
- Published
- 2018-10-22
- Volume
- 112
- Issue
- 6
- Pages
- 547-573
- ISSN
- 0340-1200 (ISSN-L)
- 1432-0770
- OpenAlex
- W2753555157
- Cited by
- 98
- References
- 78
- Citations by year
- 2026: 5 · 2025: 9 · 2024: 9 · 2023: 14 · 2022: 7 · 2021: 31 · 2020: 18 · 2019: 4 · 2018: 1
- FWCI
- 7.97
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- www.springer.com/tdm (tdm)
- Locations
- Biological Cybernetics (publishedVersion)
- PubMed (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Life Sciences › Neuroscience › Cognitive Neuroscience › Embodied and Extended Cognition
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Computer Science › Artificial Intelligence › Gaussian Processes and Bayesian Inference
- Keywords
- free energy principle, variational inference, deep generative models, latent variable models, evolution strategies, homeostasis, goal-directed behavior, recurrent neural networks, mountain car problem, active sampling
- MeSH terms
- Algorithms, Cognition, Environment, Goals, Humans, Models, Theoretical, Perception, Neural Networks, Computer, Nonlinear Dynamics
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Deep active inference. https://app.cochat.ai/papers/10.1007/s00422-018-0785-7
- The paper
- Ueltzhöffer, K. (2018). Deep active inference. Biological Cybernetics. https://doi.org/10.1007/s00422-018-0785-7
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.