Bayes in the Brain—On Bayesian Modelling in Neuroscience
The DOI matches its Crossref record; 68 of the 68 references it cites were found in a registry or an index.
Abstract
According to a growing trend in theoretical neuroscience, the human perceptual system is akin to a Bayesian machine. The aim of this article is to clearly articulate the claims that perception can be considered Bayesian inference and that the brain can be considered a Bayesian machine, some of the epistemological challenges to these claims; and some of the implications of these claims. We address two questions: (i) How are Bayesian models used in theoretical neuroscience? (ii) From the use of Bayesian models in theoretical neuroscience, have we learned or can we hope to learn that perception is Bayesian inference or that the brain is a Bayesian machine? From actual practice in theoretical neuroscience, we argue for three claims. First, currently Bayesian models do not provide mechanistic explanations; instead they are useful devices for predicting and systematizing observational statements about people's performances in a variety of perceptual tasks. That is, currently we should have an instrumentalist attitude towards Bayesian models in neuroscience. Second, the inference typically drawn from Bayesian behavioural performance in a variety of perceptual tasks to underlying Bayesian mechanisms should be understood within the three-level framework laid out by David Marr ([1982]). Third, we can hope to learn that perception is Bayesian inference or that the brain is a Bayesian machine to the extent that Bayesian models will prove successful in yielding secure and informative predictions of both subjects' perceptual performance and features of the underlying neural mechanisms. 1 Introduction2 Theoretical Neuroscientists meet Bayes3 Is Perception Bayesian Inference?4 How Should we Understand the Inference from Bayesian Observers to Bayesian Brains?5 How Could we Discover that Brains are Bayesian?6 Conclusion
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“However, the meaningfulness of a Bayes-optimal result is questionable [48] and the use of Bayesian modelling is disputed [49–53].”
From A model of cue integration as vector summation in the insect brain (2023) · cites this paper as [53]
In the citing paper
“The Bayesian brain hypothesis frequently operates in an “as-if” mode of explanation: the brain acts as if it were performing Bayesian inference, without claiming that actual Bayesian computations are implemented in neural tissue (Colombo and Seriès 2012; Danks 2014).”
From The myth of the Bayesian brain (2025) · cites this paper as Colombo and Seriès 2012
In the citing paper
“Colombo and Seriès (2012) further noted that many such models serve primarily as descriptive tools and lack correspondence with identifiable neural mechanisms, making the inference from Bayesian behavior to Bayesian brains highly tenuous.”
From The myth of the Bayesian brain (2025) · cites this paper as Colombo and Seriès 2012
Citations
Cites
Hodgkin, A. L., & Huxley, A. F. (1952) · The Journal of Physiology · doi:10.1113/jphysiol.1952.sp004764 · Cited by 23,191
Advances in neural information processing systems 7
(1996) · Computers & Mathematics with Applications · doi:10.1016/s0898-1221(96)90463-0 · Cited by 14,281
Marr, D. (1982) · Cumulative Index of Computer Aided Architectural Design · Cited by 5,733
Humans integrate visual and haptic information in a statistically optimal fashion
Ernst, M. O., & Banks, M. S. (2002) · Nature · doi:10.1038/415429a · Cited by 4,844
(2005) · Choice Reviews Online · doi:10.5860/choice.43-0626 · Cited by 4,674
Stanford encyclopedia of philosophy
(1998) · Choice Reviews Online · doi:10.5860/choice.35-4860 · Cited by 3,398
Machamer, P., Darden, L., & Craver, C. F. (2000) · Philosophy of Science · doi:10.1086/392759 · Cited by 3,179 · cited 1×
“Shimojo and Nakayama [1992]; Maloney [2002]; Fiser et al.”
BackgroundThe Bayesian brain: The role of uncertainty in neural coding and computation
Knill, D. C., & Pouget, A. (2004) · Trends in Neurosciences · doi:10.1016/j.tins.2004.10.007 · Cited by 2,588 · cited 1×
“Granted this goal of theoretically neuroscience, and granted that currently Bayesian models should be understood as no more than toolboxes for making predictions and systematizing data, how can an instrumentalist use of Bayesian models lead to gradually transforming them into mechanistic models so that a realist attit…”
BackgroundNatural Image Statistics and Neural Representation
Simoncelli, E. P., & Olshausen, B. A. (2001) · Annual Review of Neuroscience · doi:10.1146/annurev.neuro.24.1.1193 · Cited by 2,460
Treatise on Physiological Optics
von Helmholtz, H., & Southall, P. C. (1962) · Medical Entomology and Zoology · Cited by 2,196
Inference to the Best Explanation
Lipton, P. (2003) · Routledge eBooks · doi:10.4324/9780203470855 · Cited by 2,130
Bayesian integration in sensorimotor learning
Körding, K. P., & Wolpert, D. M. (2004) · Nature · doi:10.1038/nature02169 · Cited by 2,004
Scientific Realism: How Science Tracks Truth
Psillos, S. (1999) · CERN Document Server (European Organization for Nuclear Research) · Cited by 1,704
Bayesian inference with probabilistic population codes
Ma, W. J., Beck, J. M., Latham, P. E., & Pouget, A. (2006) · Nature Neuroscience · doi:10.1038/nn1790 · Cited by 1,457
Perception as Bayesian Inference
(1996) · doi:10.1017/cbo9780511984037 · Cited by 1,425
Motion illusions as optimal percepts
Weiss, Y., Simoncelli, E. P., & Adelson, E. H. (2002) · Nature Neuroscience · doi:10.1038/nn0602-858 · Cited by 1,076
Craver, C. F. (2007) · doi:10.1093/acprof:oso/9780199299317.001.0001 · Cited by 985
Mathematics, Matter and Method: Philosophical Papers
Putnam, H. (1979) · Cambridge University Press eBooks · doi:10.1017/cbo9780511625268 · Cited by 872
Noise characteristics and prior expectations in human visual speed perception
Stocker, A. A., & Simoncelli, E. P. (2006) · Nature Neuroscience · doi:10.1038/nn1669 · Cited by 797 · cited 1×
“Ma and colleagues ([2006], p.”
Friston, K. J., & Stephan, K. E. (2007) · Synthese · doi:10.1007/s11229-007-9237-y · Cited by 673
Statistically optimal perception and learning: from behavior to neural representations
Fiser, J., Berkes, P., Orbán, G., & Lengyel, M. (2010) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2010.01.003 · Cited by 672 · cited 2×
“Shimojo and Nakayama [1992]; Maloney [2002]; Fiser et al. [2010]). Alternative methods that do not require either representing or computing probabilities include: regression techniques, look-up tables, and some other supervised and unsupervised inferential strategies. Second, in the case of Bayesian modelling ‘[o]ptim…”
BackgroundProbabilistic population codes for Bayesian decision making
Beck, J. M., Ma, W. J., Kiani, R., Hanks, T., Churchland, A. K., Roitman, J., Shadlen, M. N., Latham, P. E., & Pouget, A. (2008) · Neuron · doi:10.1016/j.neuron.2008.09.021 · Cited by 645 · cited 1×
“[2006]; Beck et al. [2008]; Deneve [2008]). To carry out this project, three issues need be addressed: (i) How might neurons represent uncertainty? (ii) How might they represent probability distributions? (iii) How might they implement different approximations to Bayesian inference? Recall our case study above. Ernst…”
BackgroundVarela, F. J., & Dennett, D. C. (1993) · The American Journal of Psychology · doi:10.2307/1422870 · Cited by 629
Probabilistic models of cognition: exploring representations and inductive biases
Griffiths, T. L., Chater, N., Kemp, C., Perfors, A., & Tenenbaum, J. B. (2010) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2010.05.004 · Cited by 568
Brainard, D. H., & Freeman, W. T. (1997) · Journal of the Optical Society of America A · doi:10.1364/josaa.14.001393 · Cited by 521 · cited 2×
“on colour perception (Brainard and Freeman [1997]), motion perception (Stocker and Simoncelli [2006]), visual illusions (Weiss et al. [2002]), and sensory-motor learning (Körding and Wolpert [2004a])—would be evidence that human observers are Bayes’ optimal (Knill and Pouget [2004]).”
BackgroundBayesian Brain: Probabilistic Approaches to Neural Coding
(2006) · The MIT Press eBooks · doi:10.7551/mitpress/9780262042383.001.0001 · Cited by 511 · cited 2×
“[2002]; Doya et al. [2007]): ‘One striking observation from this work is the myriad ways in which human observers behave as optimal Bayesian observers’ (Knill and Pouget [2004], p.”
Background(2020) · The MIT Press eBooks · doi:10.7551/mitpress/11442.001.0001 · Cited by 405
Do humans optimally integrate stereo and texture information for judgments of surface slant?
Knill, D. C., & Saunders, J. A. (2003) · Vision Research · doi:10.1016/s0042-6989(03)00458-9 · Cited by 404 · cited 1×
“and systematizing data, how can an instrumentalist use of Bayesian models lead to gradually transforming them into mechanistic models so that a realist attitude towards such models can be justified? A growing number of theoretical studies have started to explore how neural mechanisms could implement the types of Bayes…”
BackgroundProbabilistic Models of the Brain : Perception and Neural Function
Rao, R. P. N., Olshausen, B. A., & Lewicki, M. S. (2002) · Cited by 385
Experience can change the 'light-from-above' prior
Adams, W. J., Graf, E. W., & Ernst, M. O. (2004) · Nature Neuroscience · doi:10.1038/nn1312 · Cited by 377 · cited 1×
“Adams et al. [2004]), currently Bayesian models are not used in this way.”
MethodExperiencing and Perceiving Visual Surfaces
Nakayama, K., & Shimojo, S. (1992) · Science · doi:10.1126/science.1529336 · Cited by 357 · cited 1×
“It is worth noticing with Simoncelli ([2009]) two things.”
BackgroundInference to the Best Explanation, 2nd edition
Farmakis, L. (2004) · Cited by 334
Maximum-Entropy and Bayesian Methods in Science and Engineering
(1988) · Springer eBooks · doi:10.1007/978-94-009-3049-0 · Cited by 314
Bayesian Spiking Neurons I: Inference
Deneve, S. (2008) · Neural Computation · doi:10.1162/neco.2008.20.1.91 · Cited by 313 · cited 1×
“Dennett [1991]). However, no-miracle arguments are controversial (see e.g. Lipton [2004], Ch. 11). And, more importantly, current practice in theoretical neuroscience––we have argued––shows that it is premature to endorse a realist attitude towards Bayesian models of sensory perception. Bayesian models do not, and do…”
BackgroundPrediction Versus Accommodation and the Risk of Overfitting
Hitchcock, C., & Sober, E. (2004) · The British Journal for the Philosophy of Science · doi:10.1093/bjps/55.1.1 · Cited by 300
Schiltz, C. (2010) · Journal of Vision · doi:10.1167/10.2.25 · Cited by 280
Logical versus Historical Theories of Confirmation
MUSGRAVE, A. (1974) · The British Journal for the Philosophy of Science · doi:10.1093/bjps/25.1.1 · Cited by 278 · cited 1×
“Theoretical neuroscientists have been increasingly using Bayesian modelling to address questions about biological perception (Rao et al. [2002]; Doya et al.”
BackgroundInference to the Best Explanation
Lipton, P. (2015) · The Routledge Companion to Philosophy of Science · doi:10.4324/9780203744857.ch21 · Cited by 266
Bayesian Computation in Recurrent Neural Circuits
Rao, R. P. N. (2004) · Neural Computation · doi:10.1162/08997660460733976 · Cited by 252
The Routledge Companion to Philosophy of Science
(2008) · Routledge eBooks · doi:10.4324/9780203000502 · Cited by 244
The loss function of sensorimotor learning
Körding, K. P., & Wolpert, D. M. (2004) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.0308394101 · Cited by 191
Bayesian decision theory as a model of human visual perception: Testing Bayesian transfer
MALONEY, L. T., & MAMASSIAN, P. (2009) · Visual Neuroscience · doi:10.1017/s0952523808080905 · Cited by 180
Regularities and causality; generalizations and causal explanations
Bogen, J. (2005) · Studies in History and Philosophy of Science Part C: Studies in History and Philosophy of Biological and Biomedical Sciences · doi:10.1016/j.shpsc.2005.03.009 · Cited by 166 · cited 1×
“4 Bogen ([2005]) was the first, or one of the first, to argue that the Hodgkin and Huxley model describes regularities but has no explanatory force.”
BackgroundProbabilistic Models of the Brain: Perception and Neural Function
(2002) · The MIT Press eBooks · doi:10.7551/mitpress/5583.001.0001 · Cited by 160
Is the Homunculus “Aware” of Sensory Adaptation?
Seriès, P., Stocker, A. A., & Simoncelli, E. P. (2009) · Neural Computation · doi:10.1162/neco.2009.09-08-869 · Cited by 156
Self-citationRapidly learned stimulus expectations alter perception of motion
Chalk, M., Seitz, A. R., & Series, P. (2010) · Journal of Vision · doi:10.1167/10.8.2 · Cited by 122 · cited 1×
“and Wolpert [2004b]; Stocker and Simoncelli [2006]; Chalk et al. [2010])”
BackgroundSelf-citationHow Does the Brain Do Plausible Reasoning?
Jaynes, E. T. (1988) · Springer eBooks · doi:10.1007/978-94-009-3049-0_1 · Cited by 118
Perception and the Physical World: Psychological and Philosophical Issues in Perception
Heyer, D., & Mausfeld, R. J. (2002) · Medical Entomology and Zoology · Cited by 113
Spiking networks for Bayesian inference and choice
Ma, W. J., Beck, J. M., & Pouget, A. (2008) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2008.07.004 · Cited by 113 · cited 3×
“Let’s illustrate the logic underlying such a transformation by considering Ma et al.’s ([2006]) work. They tackled the questions of how neuronal activity can encode probability distributions and perform Bayesian inference by building a model network. Their methodology is top-down: the computational problem that motiva…”
BackgroundMethodStatistical Decision Theory and Biological Vision
Maloney, L. T. (2002) · Perception and the Physical World · doi:10.1002/0470013427.ch6 · Cited by 110 · cited 2×
“Musgrave ([1974]) distinguishes three ways in which predictions match novel phenomena.”
BackgroundKnill, D. C. (2005) · Journal of Vision · doi:10.1167/5.2.2 · Cited by 110 · cited 3×
“Knill and Pouget ([2004], p. 712) claim that there are ‘myriad ways in which humans behave as optimal Bayesian observers.’ But it may be the case that the “myriad ways” are in fact instances of the same type. It may be the case that, though the tasks where people behave as optimal Bayesian observers seem to be differe…”
BackgroundInfluentialBrayanov, J. B., & Smith, M. A. (2010) · Journal of Neurophysiology · doi:10.1152/jn.00814.2009 · Cited by 100 · cited 1×
“[2009]; Brayanov and Smith [2010]), are thus particularly informative for two reasons: they can lead to questioning the computational goal of the system and they can shed light into the constraints on the system at the implementational level.”
BackgroundComputation in Physical Systems
Piccinini, G., & Maley, C. J. (2010) · Cited by 96
The effect of normal aging on closed contour shape discrimination
McKendrick, A. M. (2010) · Journal of Vision · doi:10.1167/10.2.1 · Cited by 72
Ambiguity and invariance: two fundamental challenges for visual processing
Rust, N. C., & Stocker, A. A. (2010) · Current Opinion in Neurobiology · doi:10.1016/j.conb.2010.04.013 · Cited by 63 · cited 1×
“constrained to the point where it can make quantitative experimentally testable predictions’ (Stocker and Simoncelli [2006], p.”
BackgroundHammett, S. T., Champion, R. A., Thompson, P., & Morland, A. B. (2006) · Vision Research · doi:10.1016/j.visres.2006.08.013 · Cited by 61
Introduction: a Bayesian formulation of visual perception
Knill, D. C., Kersten, D., & Yuille, A. (1996) · Cambridge University Press eBooks · Cited by 59
Eckstein, M. P., Abbey, C. K., Pham, B. T., & Shimozaki, S. S. (2004) · Journal of Vision · doi:10.1167/4.12.3 · Cited by 52
Comparing Bayesian models for multisensory cue combination without mandatory integration
Beierholm, U., Shams, L. B., Kording, K. P., & Ma, W. J. (2008) · CaltechAUTHORS (California Institute of Technology) · Cited by 48
Optimal Estimation in Sensory Systems
Simoncelli, E. P. (2009) · The Cognitive Neurosciences · doi:10.7551/mitpress/8029.003.0046 · Cited by 46
Pattern Inference Theory: A Probabilistic Approach to Vision
Kersten, D., & Schrater, P. (2002) · Wiley eBooks · doi:10.1002/0470013427.ch7 · Cited by 34 · cited 1×
“This case study will help us to do three things: first, to make clearer why and how Bayesian models are used in theoretical neuroscience; second, to explain in what sense there is evidence underwriting the idea that ‘human observers behave as optimal Bayesian observers’ (Knill and Pouget [2004], p.”
BackgroundDubois, M.-C. (2003) · Cited by 30
Rational analyses, instrumentalism, and implementations
Danks, D. J. (2008) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780199216093.003.0003 · Cited by 29 · cited 1×
“neural hardware allows theoretical neuroscientists to be able to use Bayesian models to generate predictions and systematize data, while remaining agnostic about the underlying mechanism (on instrumentalism and Marr’s framework, see Danks [2008]).”
BackgroundLong-Range Out-of-Sample Properties of Autoregressive Neural Networks
Leoni, P. L. (2008) · Neural Computation · doi:10.1162/neco.2009.09-07-615 · Cited by 22
Vision, Psychophysics and Bayes
Schrater, P., & Kersten, D. (2002) · The MIT Press eBooks · doi:10.7551/mitpress/5583.003.0006 · Cited by 10
Devitt, M. (2015) · Routledge eBooks · doi:10.4324/9780203744857.ch24 · Cited by 7
Research works printed but not linked: 4
- Adams W. J., Prior, 8
- Devitt M., S. Psillos and M. Curd (eds), 35
- Maloney L. T., Perception and the Physical World: Psychological and Philosophical Issues in Perception, 89
- Schrater P., Statistical Theories of the Brain
Cited by
The Border Between Seeing and Thinking
Block, N. (2023) · doi:10.1093/oso/9780197622223.001.0001 · Cited by 367
A predictive coding perspective on autism spectrum disorders
van Boxtel, J. J. A., & Lu, H. (2013) · Frontiers in Psychology · doi:10.3389/fpsyg.2013.00019 · Cited by 210
“Most prominently, the Bayesian account lacks connections to the neural computation performed by the brain, and does not provide mechanistic explanations for ASD (Rust and Stocker, 2010; Colombo and Series, 2012).”
BackgroundLearning what to expect (in visual perception)
Seriès, P., & Seitz, A. R. (2013) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2013.00668 · Cited by 194
“In parallel to the experimental study of expectations, a growing body of theoretical work suggests that visual perception is akin to Bayesian Inference (e.g. Knill and Pouget 2004; Fiser et al, 2010; Colombo and Seriès, 2012; Friston, 2012).”
BackgroundInfluentialSelf-citationCognitive Processes in Decisions Under Risk are not the Same as in Decisions Under Uncertainty
Volz, K. G., & Gigerenzer, G. (2012) · Frontiers in Neuroscience · doi:10.3389/fnins.2012.00105 · Cited by 152
“In the words of Colombo and Seriès (2012): “that the brain is a Bayesian machine does not follow from the fact that Bayesian models are used to study the brain and the behavior it generates” (p. 2).”
Predictive Processing and the Representation Wars
Williams, D. (2018) · Minds and Machines · doi:10.1007/s11023-017-9441-6 · Cited by 134
Metadata
- Authors
- Colombo, M., & Seriès, P.
- Affiliations
- University of Edinburgh (GB)
- Venue
- The British Journal for the Philosophy of Science
- Publisher
- University of Chicago Press
- Published
- 2012-09-01
- Volume
- 63
- Issue
- 3
- Pages
- 697-723
- ISSN
- 0007-0882 (ISSN-L)
- 1464-3537
- OpenAlex
- W2117986074
- Cited by
- 164
- References
- 68
- Citations by year
- 2026: 11 · 2025: 26 · 2024: 5 · 2023: 44 · 2022: 12 · 2021: 11 · 2020: 7 · 2019: 9 · 2018: 7 · 2017: 9 · 2016: 5 · 2015: 5 · 2014: 3 · 2013: 7 · 2012: 3
- FWCI
- 38.81
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- The British Journal for the Philosophy of Science (publishedVersion)
- Research portal (Tilburg University) (publishedVersion)
- Data Archiving and Networked Services (DANS) (submittedVersion)
- HighWire Press Open Archive (submittedVersion)
- Funders
- British Society for the Philosophy of Science
- Engineering and Physical Sciences Research Council
- Topics
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Philosophy of Biology and Scientific Explanation
- Life Sciences › Neuroscience › Cognitive Neuroscience › Embodied and Extended Cognition
- Social Sciences › Arts and Humanities › Philosophy › Philosophy and Theoretical Science
- Keywords
- Bayesian inference, perception, Bayesian brain, theoretical neuroscience, instrumentalism, mechanistic explanation, epistemology of neuroscience, perceptual performance, neural mechanisms, prediction
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
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.