CoChat
Sign Up
Cited by 535Neural dynamics and brain functionNeural and Behavioral Psychology StudiesCognitive Computing and Networks

Cognitive computational neuroscience

Open access
Kriegeskorte, Nikolaus, Douglas, Pamela K.·2018·Nature Neuroscience
DOI 10.1038/s41593-018-0210-5PMC PMC6706072
Open this paper in CoChat — free account

Abstract

To learn how cognition is implemented in the brain, we must build computational models that can perform cognitive tasks, and test such models with brain and behavioral experiments. Cognitive science has developed computational models that decompose cognition into functional components. Computational neuroscience has modeled how interacting neurons can implement elementary components of cognition. It is time to assemble the pieces of the puzzle of brain computation and to better integrate these separate disciplines. Modern technologies enable us to measure and manipulate brain activity in unprecedentedly rich ways in animals and humans. However, experiments will yield theoretical insight only when employed to test brain-computational models. Here we review recent work in the intersection of cognitive science, computational neuroscience and artificial intelligence. Computational models that mimic brain information processing during perceptual, cognitive and control tasks are beginning to be developed and tested with brain and behavioral data. The authors review recent work at the intersection of cognitive science, computational neuroscience and artificial intelligence that develops and tests computational models mimicking neural and cognitive function during a wide range of tasks.

AI snapshot

Written from the paper’s full text.

TL;DR

The paper argues that understanding cognition requires task-performing computational models tested with brain and behavioral data, integrating cognitive science, computational neuroscience, and artificial intelligence into a “cognitive computational neuroscience.” It reviews bottom-up and top-down modeling approaches and calls for closer integration across Marr’s levels.

Key findings

  • Bottom-up models—connectivity/dynamics, decoding, and representational models—can constrain theory but generally do not perform the cognitive tasks in question and thus do not fully explain computational mechanisms.
  • Deep convolutional neural network models trained for object classification currently provide the best models of visual representations in inferior temporal cortex in humans and monkeys; higher layers better match later ventral-stream representations, and such models typically have millions of parameters.
  • Cognitive models—production systems, reinforcement learning (model-free, model-based, episodic control), and Bayesian models—decompose higher cognition into functional components; Bayesian nonparametric models can explain single-example concept formation, but current inference algorithms require large amounts of computation and do not yet scale to real-world challenges.
  • The brain combines statistical and computational efficiency on a 20-watt power budget in ways beyond current AI, motivating integration of Bayesian inference with neural network models.

Methods

  • Design: review/perspective article synthesizing recent work at the intersection of cognitive science, computational neuroscience, and artificial intelligence.
  • Data or sample: no primary empirical sample; the paper reviews prior computational models, brain-activity studies, and behavioral findings.
  • Main analysis: conceptual synthesis organized into “from experiment toward theory” (connectivity, decoding, representational models) and “from theory to experiment” (neural network and cognitive models), with discussion of Marr’s levels and open-science collaboration.

Limitations

  • The paper states that current neural network models abstract from neurobiology (for example, action potentials), cognitive models are often implementation-independent, relating Bayesian inference and learning to brain implementation remains challenging, and current Bayesian inference algorithms do not scale to real-world concept formation.
  • Evident from the text: as a review, it presents no new experiments or quantitative meta-analysis, and its proposed integration remains programmatic, requiring collaborations and shared components across disciplines.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41593-018-0210-5

Cites

151 of 151 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 62 cited sources checked · 0 supported · 0 flagged

  • Premisecitation

    With the advent of human functional brain imaging, scientists began to relate cognitive theories to the human brain.

    “With the advent of human functional brain imaging, scientists began to relate cognitive theories to the human brain.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [7] Gazzaniga, M. S. ed. The Cognitive Neurosciences (MIT Press, Cambridge, MA, USA, 2004).

      Not checked in the free run.

  • Premisereasoning

    Cognitive science and cognitive neuroscience parted ways in the 1990s.

    “As a result, cognitive science and cognitive neuroscience parted ways in the 1990s.”
    Quote in the paperSubstantial claim
  • Interpretationreasoning

    Task-performing computational models that explain how cognition arises from neurobiologically plausible dynamic components will be central to a new cognitive computational neuroscience.

    “Here we argue that task-performing computational models that explain how cognition arises from neurobiologically plausible dynamic components will be central to a new cognitive computational neuroscience.”
    Quote in the paperNot a substantial claim
  • Premisecitation

    Computational neuroscience has modeled how interacting neurons can implement elementary components of cognition.

    “Computational neuroscience has modeled how interacting neurons can implement elementary components of cognition.”
    • Also stated in: § Abstract · plain
      “In the past two decades, the field developed mathematical models of elementary computational components and their implementation with biological neurons27,28.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [27] Dayan, P. & Abbott, L. F. Theoretical Neuroscience: Computational and Mathematical Modeling of Neural Systems (MIT Press, Cambridge, MA, USA, 2001).

      Not checked in the free run.

    • cites [28] Abbott, L. F. Theoretical neuroscience rising. Neuron 60, 489–495 (2008). source

      Not checked in the free run.

  • Interpretationreasoning

    Experiments will yield theoretical insight only when employed to test brain-computational models.

    “However, experiments will yield theoretical insight only when employed to test brain-computational models.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Newell, A. You can’t play 20 questions with nature and win: projective comments on the papers of this symposium. Technical Report, School of Computer Science, Carnegie Mellon University (1973). source

      Not checked in the free run.

  • Classified as probably not a substantial claim. 27
  • May combine two claims. 12

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    The transition from cognitive psychology to cognitive science was defined by the introduction of task-performing computational models.

    “In the spirit of Newell’s critique, the transition from cognitive psychology to cognitive science was defined by the introduction of task-performing computational models.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Computer hardware and machine learning were not sufficiently advanced to simulate cognitive processes in their full complexity.

    “However, computer hardware and machine learning were not sufficiently advanced to simulate cognitive processes in their full complexity.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationcitation

    If we had a full understanding of an information-processing mechanism, we should be able to engineer it.

    “If we did have a full understanding of an information-processing mechanism, then we should be able to engineer it.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisecitation

    Hypothesis testing needed to be complemented by the construction of comprehensive task-performing computational models.

    “Hypothesis testing, in his view, needed to be complemented by the construction of comprehensive task-performing computational models.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Newell, A. You can’t play 20 questions with nature and win: projective comments on the papers of this symposium. Technical Report, School of Computer Science, Carnegie Mellon University (1973). source

      Not checked in the free run.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisenone given

    Early cognitive science developments relied on behavioral data alone and did not leverage constraints provided by the anatomy and activity of the brain.

    “Moreover, these early developments relied on behavioral data alone and did not leverage constraints provided by the anatomy and activity of the brain.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S1 (asserted).

    Premisecitation

    Testing verbally defined hypotheses about cognition might never lead to a computational understanding.

    “Newell argued that testing verbally defined hypotheses about cognition might never lead to a computational understanding.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [1] Newell, A. You can’t play 20 questions with nature and win: projective comments on the papers of this symposium. Technical Report, School of Computer Science, Carnegie Mellon University (1973). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    A brain-computational model of face recognition will have to explain the spatial clusters of face-selective units and the selectivities and invariances observed with fMRI and invasive recordings.

    “A brain-computational model of face recognition16 will have to explain the spatial clusters of face-selective units and the selectivities and invariances observed with fMRI17,18 and invasive recordings14,19.”
    Quote in the paperSubstantial claim0 of 5 sources checked
    • cites [16] Yildirim, I. et al. Efficient and robust analysis-by-synthesis in vision: a computational framework, behavioral tests, and modeling neuronal representations. in Annual Conference of the Cognitive Science Society (eds. Noelle, D. C. et al.) (Cognitive Science Society, Austin, TX, USA, 2015).

      Not checked in the free run.

    • cites [17] Kriegeskorte, N., Formisano, E., Sorger, B. & Goebel, R. Individual faces elicit distinct response patterns in human anterior temporal cortex. Proc. Natl Acad. Sci. USA 104, 20600–20605 (2007). source

      Not checked in the free run.

    • cites [18] Anzellotti, S., Fairhall, S. L. & Caramazza, A. Decoding representations of face identity that are tolerant to rotation. Cereb. Cortex 24, 1988–1995 (2014). source

      Not checked in the free run.

    • cites [14] Freiwald, W. A. & Tsao, D. Y. Functional compartmentalization and viewpoint generalization within the macaque face-processing system. Science 330, 845–851 (2010). source

      Not checked in the free run.

    • cites [19] Chang, L. & Tsao, D. Y. The code for facial identity in the primate brain. Cell 169, 1013–1028.e14 (2017). source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Primate face patches revealed high densities of face-selective neurons, with invariances emerging at higher stages of hierarchical processing.

    “Localized with fMRI and probed with invasive electrode recordings, the primate face patches revealed high densities of face-selective neurons13, with invariances emerging at higher stages of hierarchical processing, including mirror-symmetric tuning and view-tolerant representations of individual faces in the anteriormost patch14.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [13] Tsao, D. Y., Freiwald, W. A., Tootell, R. B. & Livingstone, M. S. A cortical region consisting entirely of face-selective cells. Science 311, 670–674 (2006). source

      Not checked in the free run.

    • cites [14] Freiwald, W. A. & Tsao, D. Y. Functional compartmentalization and viewpoint generalization within the macaque face-processing system. Science 330, 845–851 (2010). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationreasoning

    Deep neural network models' functionality is inspired by brains and could be implemented with biological neurons.

    “However, their functionality is inspired by brains and could be implemented with biological neurons.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Recent advances in machine learning, boosted by growing computational power and larger datasets, have brought progress at perceptual, cognitive and control challenges.

    “Recent advances in machine learning, boosted by growing computational power and larger datasets from which to learn, have brought progress at perceptual39, cognitive40 and control challenges41.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [39] Krizhevsky, A., Sutskever, I. & Hinton, G. E. ImageNet classification with deep convolutional neural networks. in Advances in Neural Information Processing Systems 25 1097–1105 (Curran Associates, Red Hook, NY, USA, 2012).

      Not checked in the free run.

    • cites [40] Silver, D. et al. Mastering the game of Go with deep neural networks and tree search. Nature 529, 484–489 (2016). source

      Not checked in the free run.

    • cites [41] Mnih, V. et al. Human-level control through deep reinforcement learning. Nature 518, 529–533 (2015). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    If computational models are to explain animal and human cognition, they will have to perform feats of intelligence.

    “If computational models are to explain animal and human cognition, they will have to perform feats of intelligence.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The face perception literature shows solid progress in mapping the anatomical substrate and characterizing neuronal responses but lacks definitive computational models.

    “The example of face perception illustrates, on one hand, the solid progress in mapping the anatomical substrate and characterizing neuronal responses15 and, on the other, the lack of definitive computational models.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15] Grill-Spector, K., Weiner, K. S., Kay, K. & Gomez, J. The functional neuroanatomy of human face perception. Annu. Rev. Vis. Sci. 3, 167–196 (2017). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisenone given

    Deep neural network models employ only a small subset of the dynamic capabilities of biological neurons, abstracting from fundamental features such as action potentials.

    “These models employ only a small subset of the dynamic capabilities of biological neurons, abstracting from fundamental features such as action potentials.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given

    Most component functions in computational neuroscience are computationally simple but provide building blocks for cognition.

    “Most of these component functions are computationally simple, but they provide building blocks for cognition.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisereasoning

    The anatomical connectivity matrix typically resembles the functional connectivity matrix because regions interact through anatomical pathways.

    “The anatomical connectivity matrix typically resembles the functional connectivity matrix because regions interact through anatomical pathways.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisenone given

    Connectivity models go beyond the localization of activated regions and characterize the interactions between regions.

    “Connectivity models go beyond the localization of activated regions and characterize the interactions between regions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Information exchange incurs costs that scale with the distance between communicating regions-costs in physical connections, energy and signal latency.

    “After all, information exchange incurs costs that scale with the distance between the communicating regions—costs in terms of physical connections, energy and signal latency.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Connectivity models, decoding models, and representational models provide constraints for computational theory but do not in general perform cognitive tasks and thus fall short of explaining the computational mechanism underlying task performance.

    “The models employed in this literature provide constraints for computational theory, but they do not in general perform the cognitive tasks in question and thus fall short of explaining the computational mechanism underlying task performance.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisenone given

    Cognitive models decompose cognition into meaningful functional components.

    “Cognitive models, including the three classes highlighted here, decompose cognition into meaningful functional components.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Premisereasoning

    It is always in the context of prior theory (generative models of the observed processes) that better observations advance our understanding.

    “However, it is always in the context of prior theory (generative models of the observed processes) that better observations advance our understanding.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    When particular content is decodable from activity in a brain region, this indicates the presence of the information.

    “When particular content is decodable from activity in a brain region, this indicates the presence of the information.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    Deep neural network models provide a biologically plausible account of the rapid recognition of the elements of the visual experience.

    “Deep neural network models provide a biologically plausible account of the rapid recognition of the elements of the visual experience (people, steps, oblique upward motion, handrail).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Marr (1982) offered a distinction of three levels of analysis: computational theory, representation and algorithm, and neurobiological implementation.

    “Marr (1982) offered a distinction of three levels of analysis: (i) computational theory, (ii) representation and algorithm, and (iii) neurobiological implementation115.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [115] Marr, D. Vision: A Computational Investigation into the Human Representation and Processing of Visual Information (MIT Press, Cambridge, MA, USA, 2010). source

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning

    Deep neural network models cannot yet explain how a child understands relationships among elements, physical interactions, goals, function of an escalator, or how she can imagine the experience and instantly form a new concept.

    “However, they cannot explain yet how the child understands the relationships among the elements, the physical interactions of the objects, the people’s goal to go up, and the function of the escalator, or how she can imagine the experience and instantly form a new concept.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    AI builds representations and algorithms that combine simple components to implement complex feats of intelligence.

    “AI builds representations and algorithms that combine simple components to implement complex feats of intelligence.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    It may not be possible to understand cognition without considering its implementation in the brain or to make sense of neuronal circuits except in the context of the cognitive functions they support.

    “It may therefore not be possible to understand cognition without considering its implementation in the brain or, conversely, to make sense of neuronal circuits except in the context of the cognitive functions they support.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Brains are the product of evolution and development, processes that are not constrained to generate systems whose behavior can be perfectly captured at some abstract level of description.

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisereasoning

    Current inference algorithms require large amounts of computation and do not yet scale to real-world challenges such as forming the new concept "escalator" from a single visual experience.

    “However, current inference algorithms require large amounts of computation and, as a result, do not yet scale to real-world challenges such as forming the new concept “escalator” from a single visual experience.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisenone given

    Computational neuroscience proceeds from the bottom up, composing neuronal building blocks into representations and algorithms thought to be useful components in the context of the brain's overall function.

    “Computational neuroscience proceeds from the bottom up, composing neuronal building blocks into representations and algorithms thought to be useful components in the context of the brain’s overall function.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisenone given

    Cognitive science starts from computational theory, decomposing cognition into components and developing representations and algorithms from the top down.

    “Cognitive science starts from computational theory, decomposing cognition into components and developing representations and algorithms from the top down.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Cognitive models are essential for cognitive computational neuroscience because they enable us to see the whole as we attempt to understand the roles of the parts.

    “Cognitive models are essential for cognitive computational neuroscience because they enable us to see the whole as we attempt to understand the roles of the parts.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    The ability to decode particular information does not amount to a full account of the neuronal code: it does not specify the representational format or what other information might additionally be present.

    “However, the ability to decode particular information does not amount to a full account of the neuronal code: it doesn’t specify the representational format (beyond linear decodability) or what other information might additionally be present.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisereasoning

    This enables engineers to abstract from the circuits when designing algorithms.

    “This enables the engineers to abstract from the circuits when designing the algorithms.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisereasoning

    Causal models of processes in the world can give an agent a deeper understanding and thus a better basis for inferences and actions.

    “In particular, causal models of processes in the world (how objects cause images, how the present causes the future) can give an agent a deeper understanding and thus a better basis for inferences and actions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationreasoning

    Bayesian nonparametric models explain how deep inferences and concept formation from single experiences are even possible.

    “Bayesian nonparametric models explain how deep inferences and concept formation from single experiences are even possible.”
    Quote in the paperNot a substantial claim
34 of 34

  • The paper relies on this without support.

    Modern technologies enable us to measure and manipulate brain activity in unprecedentedly rich ways in animals and humans.

    “Modern technologies enable us to measure and manipulate brain activity in unprecedentedly rich ways in animals and humans.”

  • The paper relies on this without support.

    Biological organisms have limited time to learn and limited memory, and must avoid interactions that might kill them.

    “Biological organisms, however, have limited time to learn and limited memory, and must avoid interactions that might kill them.”

  • The paper relies on this without support.

    The brain seamlessly merges bottom-up discriminative and top-down generative computations in perceptual inference, and model-free and model-based control.

    “The brain seamlessly merges bottom-up discriminative and top-down generative computations in perceptual inference, and model-free and model-based control.”

  • The paper relies on this without support.

    Brains are the product of evolution and development, processes that are not constrained to generate systems whose behavior can be perfectly captured at some abstract level of description.

    “Brains are the product of evolution and development, processes that are not constrained to generate systems whose behavior can be perfectly captured at some abstract level of description.”

  • The paper relies on this without support.

    On a 20-watt power budget, the brain’s algorithms combine statistical and computational efficiency in ways that are beyond current AI of either the Bayesian or the neural network variety.

    “On a 20-watt power budget, the brain’s algorithms combine statistical and computational efficiency in ways that are beyond current AI of either the Bayesian or the neural network variety.”

  • The paper relies on this without support.

    Recent advances in computer hardware and software enable AI-scale modeling of the mind and brain.

    “Recent advances in computer hardware and software enable AI-scale modeling of the mind and brain.”
6 of 6

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 mathematical formalism of Bayesian approaches has particular appeal in an era when neuroscience increasingly values computational sophistication.

    “The mathematical formalism of Bayesian approaches has particular appeal in an era when neuroscience increasingly values computational sophistication (Kriegeskorte and Douglas 2018).”

    From The myth of the Bayesian brain (2025) · cites this paper as Kriegeskorte and Douglas 2018

  • Formal artificial intelligence models provide flexibility in planning and account for a range of biological processes at neural and behavioural levels.

    “Formal artificial intelligence models provide flexibility in planning and account for a range of biological processes at neural and behavioural levels 204 .”

    From From cognitive maps to spatial schemas (2023) · cites this paper as [204]

  • Navigation planning follows a hierarchical representation of space that consists of a series of contexts that, in humans, is mediated by the dorsomedial PFC.

    “Studies of navigation support the idea that navigation planning follows a hierarchical representation of space that consist of a series of contexts that, in humans, is mediated by the dorsomedial PFC 128,133,159,204 .”

    From From cognitive maps to spatial schemas (2023) · cites this paper as [204]

Cites

Flagged first, then by citations

Research works printed but not linked: 5

  1. Anderson, J. R. The Architecture of Cognition (Harvard Univ. Press, Cambridge, MA, USA, 1983).
  2. Gazzaniga, M. S. ed. The Cognitive Neurosciences (MIT Press, Cambridge, MA, USA, 2004).
  3. Yildirim, I. et al. Efficient and robust analysis-by-synthesis in vision: a computational framework, behavioral tests, and modeling neuronal representations. in Annual Conference of the Cognitive Science Society (eds. Noelle, D. C. et al.) (Cognitive Science Society, Austin, TX, USA, 2015).
  4. Rezende, D., Mohamed, S., Danihelka, I., Gregor, K. & Wierstra, D. One-shot generalization in deep generative models. Proc. Int. Conf. Mach. Learn. Appl. 48, 1521–1529 (2016).
  5. Kandel, E. R., Schwartz, J. H., Jessell, T. M., Siegelbaum, S. A. & Hudspeth, A. J. Principles of Neural Science (McGraw-Hill Professional, New York, 2013).

Cited by

Sorted by citations

Related

Authors
Kriegeskorte, N., & Douglas, P. K.
Affiliations
  • Mortimer B. Zuckerman Mind Brain Behavior Institute (US)
  • University of California, Los Angeles (US)
Corresponding author
Kriegeskorte, Nikolaus
ORCID
Venue
Nature Neuroscience
Publisher
Springer Science and Business Media LLC
Published
2018-08-20
Volume
21
Issue
9
Pages
1148-1160
ISSN
  • 1097-6256 (ISSN-L)
  • 1546-1726
DOI
10.1038/s41593-018-0210-5
PMC
PMC6706072
OpenAlex
W2887972576
Semantic Scholar
fe65133015723d49f8090e0aec099e7b61b8a54b
Cited by
535
References
151
Citations by year
2026: 54 · 2025: 69 · 2024: 89 · 2023: 69 · 2022: 100 · 2021: 62 · 2020: 53 · 2019: 37 · 2018: 2
FWCI
5.16
Type
journal article
Language
en
Open access
Licence
www.springer.com/tdm (tdm)
Locations
  • Nature Neuroscience (publishedVersion)
  • PubMed (publishedVersion)
  • PubMed Central (acceptedVersion)
Funders
National Institute of Mental Health (K01MH110645)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural dynamics and brain function
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neural and Behavioral Psychology Studies
  • Physical Sciences › Computer Science › Artificial Intelligence › Cognitive Computing and Networks
Keywords
cognitive modeling, artificial intelligence, neural information processing, perceptual tasks, cognitive control
MeSH terms
Animals, Cognition, Computer Simulation, Humans, Models, Neurological, Neurosciences, Computational Biology

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.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

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