Compositional diversity in visual concept learning DOI Creative Commons
Yanli Zhou,

Reuben Feinman,

Brenden M. Lake

et al.

Cognition, Journal Year: 2024, Volume and Issue: 244, P. 105711 - 105711

Published: Jan. 14, 2024

Language: Английский

Replay and compositional computation DOI Creative Commons
Zeb Kurth‐Nelson, Timothy E.J. Behrens,

Greg Wayne

et al.

Neuron, Journal Year: 2023, Volume and Issue: 111(4), P. 454 - 469

Published: Jan. 13, 2023

Language: Английский

Citations

60

Unpacking the Complexities of Consciousness: Theories and Reflections DOI Creative Commons
Liad Mudrik, Mélanie Boly,

Stanislas Dehaene

et al.

Neuroscience & Biobehavioral Reviews, Journal Year: 2025, Volume and Issue: unknown, P. 106053 - 106053

Published: Feb. 1, 2025

As the field of consciousness science matures, research agenda has expanded from an initial focus on neural correlates consciousness, to developing and testing theories consciousness. Several have been put forward, each aiming elucidate relationship between brain function. However, there is ongoing, intense debate regarding whether these examine same phenomenon. And, despite ongoing efforts, it seems like so far failed converge around any single theory, instead exhibits significant polarization. To advance this discussion, proponents five prominent consciousness-Global Neuronal Workspace Theory (GNWT), Higher-Order Theories (HOT), Integrated Information (IIT), Recurrent Processing (RPT), Predictive (PP)-engaged in a public 2022, as part annual meeting Association for Scientific Study Consciousness (ASSC). They were invited clarify explananda their theories, articulate core mechanisms underpinning corresponding explanations, outline foundational premises. This was followed by open discussion that delved into testability potential evidence could refute them, areas consensus disagreement. Most importantly, demonstrated at stage, more controversy than agreement pertaining most basic questions what is, how identify conscious states, required theory Addressing crucial advancing towards deeper understanding comparison competing theories.

Language: Английский

Citations

3

A language of thought for the mental representation of geometric shapes DOI
Mathias Sablé-Meyer, Kevin Ellis,

Josh Tenenbaum

et al.

Cognitive Psychology, Journal Year: 2022, Volume and Issue: 139, P. 101527 - 101527

Published: Nov. 17, 2022

Language: Английский

Citations

51

Spatiotemporally distributed frontotemporal networks for sentence reading DOI Creative Commons
Oscar Woolnough, Cristian Donos, Elliot Murphy

et al.

Proceedings of the National Academy of Sciences, Journal Year: 2023, Volume and Issue: 120(17)

Published: April 17, 2023

Reading a sentence entails integrating the meanings of individual words to infer more complex, higher-order meaning. This highly rapid and complex human behavior is known engage inferior frontal gyrus (IFG) middle temporal (MTG) in language-dominant hemisphere, yet whether there are distinct contributions these regions reading still unclear. To probe neural spatiotemporal dynamics, we used direct intracranial recordings measure activity while sentences, meaning-deficient Jabberwocky lists or pseudowords. We isolated two functionally spatiotemporally frontotemporal networks, each sensitive aspects word composition. The first distributed network engages IFG MTG, with preceding MTG. Activity this ramps up over duration reduced absent during lists, implying its role derivation sentence-level second superior IFG, responses leading those lobe, shows greater activation for list than suggesting that sentential context enables efficiency lexical and/or phonological processing words. These adjacent, dissociable mechanisms word- processes shed light on richly layered semantic networks enable us fluently read. results imply distributed, dynamic computation across language rather clear dichotomy between structures.

Language: Английский

Citations

33

Continual task learning in natural and artificial agents DOI Creative Commons
Timo Flesch, Andrew Saxe, Christopher Summerfield

et al.

Trends in Neurosciences, Journal Year: 2023, Volume and Issue: 46(3), P. 199 - 210

Published: Jan. 20, 2023

How do humans and other animals learn new tasks? A wave of brain recording studies has investigated how neural representations change during task learning, with a focus on tasks can be acquired coded in ways that minimise mutual interference. We review recent work explored the geometry dimensionality neocortex, computational models have exploited these findings to understand may partition knowledge between tasks. discuss ideas from machine including those combine supervised unsupervised are helping neuroscientists natural learned biological brains.

Language: Английский

Citations

28

Concepts at the Interface DOI
Nicholas Shea

Oxford University Press eBooks, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 9, 2024

Abstract Research on concepts has concentrated the way people apply online, when presented with a stimulus. Just as important, however, is use of offline, planning what to do or thinking about case. There strong evidence that inferences driven by conceptual thought draw heavily special-purpose resources: sensory, motoric, affective, and evaluative. At same time, afford general-purpose recombination support domain-general reasoning processes—phenomena have long been focus philosophers. growing consensus theory must encompass both kinds process. This book shows how are able act an interface between systems. Concept-driven can take advantage complementary costs benefits each. The lays out empirically-based account different ways in which takes us new conclusions underpins planning, decision-making, action. It also spells three useful implications account. First, it allows reconstruct commonplace idea draws meaning concept. Second, offers insight into human cognition avoids frame problem complementary, less discussed, ‘if-then problem’ for nested processing dispositions. Third, metacognition concept-driven various ways. framework developed elucidates makes especially powerful cognitive resource.

Language: Английский

Citations

15

Distributed Sensitivity to Syntax and Semantics throughout the Language Network DOI Creative Commons
Cory Shain, Hope Kean, Colton Casto

et al.

Journal of Cognitive Neuroscience, Journal Year: 2024, Volume and Issue: 36(7), P. 1427 - 1471

Published: Jan. 1, 2024

Abstract Human language is expressive because it compositional: The meaning of a sentence (semantics) can be inferred from its structure (syntax). It commonly believed that syntax and semantics are processed by distinct brain regions. Here, we revisit this claim using precision fMRI methods to capture separation or overlap function in the brains individual participants. Contrary prior claims, find distributed sensitivity both throughout broad frontotemporal network. Our results join growing body evidence for an integrated network human within which internal specialization primarily matter degree rather than kind, contrast with influential proposals advocate different areas types linguistic functions.

Language: Английский

Citations

14

Uniquely human intelligence arose from expanded information capacity DOI Open Access
Jessica F. Cantlon, Steven T. Piantadosi

Nature Reviews Psychology, Journal Year: 2024, Volume and Issue: 3(4), P. 275 - 293

Published: April 2, 2024

Language: Английский

Citations

11

Natural language syntax complies with the free-energy principle DOI Creative Commons
Elliot Murphy, Emma Holmes, Karl Friston

et al.

Synthese, Journal Year: 2024, Volume and Issue: 203(5)

Published: May 3, 2024

Abstract Natural language syntax yields an unbounded array of hierarchically structured expressions. We claim that these are used in the service active inference accord with free-energy principle (FEP). While conceptual advances alongside modelling and simulation work have attempted to connect speech segmentation linguistic communication FEP, we extend this program underlying computations responsible for generating syntactic objects. argue recently proposed principles economy design—such as “minimal search” criteria from theoretical syntax—adhere FEP. This affords a greater degree explanatory power FEP—with respect higher functions—and offers linguistics grounding first computability. mostly focus on building new principled relations between also show through sample preliminary examples how both tree-geometric depth Kolmogorov complexity estimate (recruiting Lempel–Ziv compression algorithm) can be accurately predict legal operations workspaces, directly line formulations variational free energy minimization. is motivate general design term Turing–Chomsky Compression (TCC). use TCC align concerns linguists normative account self-organization furnished by marshalling evidence psycholinguistics ground core efficient computation within inference.

Language: Английский

Citations

8

The relational bottleneck as an inductive bias for efficient abstraction DOI
Taylor W. Webb, Steven Frankland,

Awni Altabaa

et al.

Trends in Cognitive Sciences, Journal Year: 2024, Volume and Issue: 28(9), P. 829 - 843

Published: May 9, 2024

Language: Английский

Citations

8