Representational similarity analysis in neuroimaging: proxy vehicles and provisional representations

Roskies, Adina L.

doi:10.1007/s11229-021-03052-4

Cited by 13 publications

(15 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For detailed discussion of the limits and merits of this approach, see Roskies. 23 The model included predictors corresponding to both simple stimulus features and to participants' judgments of emotion content, supporting tests of A1 and A2. We performed an additional supporting test of H1 using a model-free approach that directly compared representational geometries across sensory areas without making any assumptions about representational content.…”

Section: Introductionmentioning

confidence: 59%

Visual and auditory brain areas share a representational structure that supports emotion perception

Sievers¹,

Parkinson²,

Kohler³

et al. 2021

Current Biology

View full text Add to dashboard Cite

show abstract

Section: Introductionmentioning

confidence: 59%

Visual and auditory brain areas share a representational structure that supports emotion perception

Sievers¹,

Parkinson²,

Kohler³

et al. 2021

Current Biology

View full text Add to dashboard Cite

show abstract

“…Neuroscientists can then use statistical methods to test the similarity of different RDMs generated from distinct representational spaces. This gives a quantitative sense of to what degree two different representational spaces are alike (Kriegeskorte & Kievit, 2013;Roskies, 2021).…”

Section: Representational Similarity Analysis (Rsa)mentioning

confidence: 99%

“…Encoding models instead work in the same direction as the flow of information in the brain to make comprehensive predictions about the representational space (Naselaris & Kay, 2015;Diedrichsen & Kriegeskorte, 2017). The model itself can be generated by statistical descriptions of participant behaviors, neural data obtained directly from proxy organisms like chimpanzees, or patterns of activity from ANN models trained on some sensory-perceptual task (Kriegeskorte & Douglas, 2019;Martin et al, 2018;Roskies, 2021). to some chosen metric (e.g Euclidean, correlation, etc.).…”

Section: Representational Similarity Analysis (Rsa)mentioning

confidence: 99%

“…Predictively adequate neural network models will exhibit a similar block-diagonal structure in their own RDMs. When cells are arranged according to observer similarity judgments, strong structural correlation provides evidence that the activation space implements a representational space (Kriegeskorte et al, 2008a;Roskies, 2021). RDMs can thus be interpreted as a simplified description of the geometry of a given representational space.…”

Section: Representational Similarity Analysis (Rsa)mentioning

confidence: 99%

“…The practices of MVPA will benefit from greater contact with the philosophy of science. In what follows, I consider in detail a specific type of MVPA called representational similarity analysis (RSA) (Kriegeskorte et al, 2008a;Roskies, 2021). RSA is noteworthy for its shared conceptual DNA with earlier connectionist approaches to cognition that use similarity in a semantic space as a measure of representational content (Churchland, 1998;Horgan & Tienson, 1996;Rumelhart et al, 1986).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Mapping representational mechanisms with deep neural networks

Kieval

2022

Synthese

View full text Add to dashboard Cite

The predominance of machine learning based techniques in cognitive neuroscience raises a host of philosophical and methodological concerns. Given the messiness of neural activity, modellers must make choices about how to structure their raw data to make inferences about encoded representations. This leads to a set of standard methodological assumptions about when abstraction is appropriate in neuroscientific practice. Yet, when made uncritically these choices threaten to bias conclusions about phenomena drawn from data. Contact between the practices of multivariate pattern analysis (MVPA) and philosophy of science can help to illuminate the conditions under which we can use artificial neural networks to better understand neural mechanisms. This paper considers a specific technique for MVPA called representational similarity analysis (RSA). I develop a theoretically-informed account of RSA that draws on early connectionist research and work on idealization in the philosophy of science. By bringing a philosophical account of cognitive modelling in conversation with RSA, this paper clarifies the practices of neuroscientists and provides a generalizable framework for using artificial neural networks to study neural mechanisms in the brain.

show abstract

Representational formats of human memory traces

et al. 2023

View full text Add to dashboard Cite

Neural representations are internal brain states that constitute the brain’s model of the external world or some of its features. In the presence of sensory input, a representation may reflect various properties of this input. When perceptual information is no longer available, the brain can still activate representations of previously experienced episodes due to the formation of memory traces. In this review, we aim at characterizing the nature of neural memory representations and how they can be assessed with cognitive neuroscience methods, mainly focusing on neuroimaging. We discuss how multivariate analysis techniques such as representational similarity analysis (RSA) and deep neural networks (DNNs) can be leveraged to gain insights into the structure of neural representations and their different representational formats. We provide several examples of recent studies which demonstrate that we are able to not only measure memory representations using RSA but are also able to investigate their multiple formats using DNNs. We demonstrate that in addition to slow generalization during consolidation, memory representations are subject to semantization already during short-term memory, by revealing a shift from visual to semantic format. In addition to perceptual and conceptual formats, we describe the impact of affective evaluations as an additional dimension of episodic memories. Overall, these studies illustrate how the analysis of neural representations may help us gain a deeper understanding of the nature of human memory.

show abstract

Representational similarity analysis in neuroimaging: proxy vehicles and provisional representations

Cited by 13 publications

References 41 publications

Visual and auditory brain areas share a representational structure that supports emotion perception

Visual and auditory brain areas share a representational structure that supports emotion perception

Mapping representational mechanisms with deep neural networks

Representational formats of human memory traces

Contact Info

Product

Resources

About