Distractor-specific control adaptation in multidimensional environments DOI Creative Commons
Davide Gheza, Wouter Kool

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Sept. 5, 2023

Abstract Goal-directed behavior requires humans to constantly manage and switch between multiple, independent, conflicting, sources of information. Conventional cognitive control tasks, however, only feature one task source distraction. Therefore, it is unclear how allocated in multidimensional environments. To address this question, we developed a multi-dimensional task-set interference paradigm, which people need distraction from three independent dimensions. We use test whether adapt prior conflict by enhancing task-relevant information or suppressing task-irrelevant Three experiments provided strong evidence for the latter hypothesis. Moreover, adaptation was highly dimension-specific. Conflict given dimension affected processing that same on subsequent trials, with no generalization. A new neural network model shows our results can be simulated when including multiple conflict-detector units. Our call an update classic models control, their neurocomputational underpinnings.

Language: Английский

Learning attentional templates for value-based decision-making DOI Creative Commons
Caroline I. Jahn, Nikola T. Markov, Britney M. Morea

et al.

Cell, Journal Year: 2024, Volume and Issue: 187(6), P. 1476 - 1489.e21

Published: Feb. 23, 2024

Attention filters sensory inputs to enhance task-relevant information. It is guided by an "attentional template" that represents the stimulus features are currently relevant. To understand how brain learns and uses templates, we trained monkeys perform a visual search task required them repeatedly learn new attentional templates. Neural recordings found templates were represented across prefrontal parietal cortex in structured manner, such perceptually neighboring had similar neural representations. When changed, template was learned incrementally shifting toward rewarded features. Finally, transformed into common value representation allowed same decision-making mechanisms deploy attention, regardless of identity template. Altogether, our results provide insight which control attention can be flexibly deployed tasks.

Language: Английский

Citations

6
Task structure tailors the geometry of neural representations in human lateral prefrontal cortex DOI Creative Commons
Apoorva Bhandari, Haley Keglovits, David Badre

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: March 7, 2024

How do human brains represent tasks of varying structure? The lateral prefrontal cortex (lPFC) flexibly represents task information. However, principles that shape lPFC representational geometry remain unsettled. We use fMRI and pattern analyses to reveal the structure geometries as humans perform two distinct categorization tasks- one with flat, conjunctive categories another hierarchical, context dependent categories. show encodes relevant information tailored intermediate dimensionality. These preferentially enhance separability variables while encoding a subset in abstract form. Specifically, flat task, global axis response abstractly, category specific local are high dimensional. In hierarchy abstractly higher level context, low dimensional, compress irrelevant encode Comparing these exposes generalizable by which tailors representations different tasks.

Language: Английский

Citations

5
Distractor-specific control adaptation in multidimensional environments DOI
Davide Gheza, Wouter Kool

Nature Human Behaviour, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 3, 2025

Language: Английский

Citations

0
Coordinated representations for naturalistic memory encoding and retrieval in hippocampal neural subspaces DOI Creative Commons
Dasom Kwon, Jungwoo Kim, Seng Bum Michael Yoo

et al.

Nature Communications, Journal Year: 2025, Volume and Issue: 16(1)

Published: Jan. 14, 2025

Our naturalistic experiences are organized into memories through multiple processes, including novelty encoding, memory formation, and retrieval. However, the neural mechanisms coordinating these processes remain elusive. Using fMRI data acquired during movie viewing subsequent narrative recall, we examine hippocampal subspaces associated with distinct characterized their relationships. We quantify in character co-occurrences valence of relationships estimate event memorability. Within hippocampus, encoding each type exhibit partial overlap, overlapping align subspace involved Notably, following boundaries, states within inversely along a shared coding axis, predicting recall performance. This novelty-memorability alignment is selectively observed but not Finally, identified functional reflect intrinsic organization hippocampus. findings offer insights how hippocampus dynamically coordinates representations underlying retrieval at population level to transform ongoing enduring memories. In this study, authors show that aligns formation retrieval, components process its organisation.

Language: Английский

Citations

0
The unity/diversity framework of executive functions: behavioral and neural evidence in older adults DOI

Sheng-Ju Guo,

Ping Wang, Lizhi Cao

et al.

GeroScience, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 31, 2025

Language: Английский

Citations

0
An image-computable model of speeded decision-making DOI Creative Commons
Paul I. Jaffe,

Gustavo X. Santiago-Reyes,

Robert J. Schafer

et al.

eLife, Journal Year: 2025, Volume and Issue: 13

Published: Feb. 28, 2025

Evidence accumulation models (EAMs) are the dominant framework for modeling response time (RT) data from speeded decision-making tasks. While providing a good quantitative description of RT in terms abstract perceptual representations, EAMs do not explain how visual system extracts these representations first place. To address this limitation, we introduce accumulator model (VAM), which convolutional neural network processing and traditional jointly fitted to trial-level RTs raw (pixel-space) stimuli individual subjects unified Bayesian framework. Models large-scale cognitive training stylized flanker task captured differences congruency effects, RTs, accuracy. We find evidence that selection task-relevant information occurs through orthogonalization relevant irrelevant demonstrating our can be used relate behavioral outputs. Together, work provides probabilistic both constraining vision with studying guide decisions.

Language: Английский

Citations

0
The role of human intraparietal sulcus in evidence accumulation revealed by EEG and model-informed fMRI DOI Creative Commons

Jaeger Wongtrakun,

Shou-Han Zhou,

Redmond G O’Connell

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 6, 2025

Abstract Sequential sampling models propose that the repeated of sensory information is a fundamental component perceptual decision-making. Electroencephalographic investigations in humans have demonstrated motor-independent representations evidence accumulation, but such observations seldom been made neuroimaging studies exploring neuroanatomical origins accumulation. Here, we aimed to reveal locus accumulation human brain by regressing an electrophysiological marker (centroparietal positivity, CPP) against changes blood oxygen level-dependent (BOLD) signal during Our cross-modal imaging approach revealed cluster within left intraparietal sulcus (IPS), located putative lateral area (region hIP3), for which BOLD signals scaled relation slope CPP. Furthermore, drift rate parameter diffusion model parametrically modulated activity overlapping region IPS. In contrast, parametric modulation reaction time distributed fronto-parietal network, demonstrating utility our isolating discrete Together, findings provide strong support involvement

Language: Английский

Citations

0
Practice Reshapes the Geometry and Dynamics of Task-tailored Representations DOI Creative Commons
Atsushi Kikumoto, Kazuhisa Shibata,

Takahiro Nishio

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 15, 2024

Extensive practice makes task performance more efficient and precise, leading to automaticity. However, theories of automaticity differ on which levels representations (e.g., low-level features, stimulus-response mappings, or high-level conjunctive memories individual events) change with practice, despite predicting the same pattern improvement power law practice). To resolve this controversy, we built recent theoretical advances in understanding computations through neural population dynamics. Specifically, hypothesized that optimizes representational geometry minimally separate highest-level contingencies needed for successful performance. This involves efficiently reaching states integrate task-critical features nonlinearly while abstracting over non-critical dimensions. test hypothesis, human participants (n = 40) engaged extensive a simple, context-dependent action selection 3 days recording EEG. During initial rapid performance, highest-level, context-specific conjunctions task-features were enhanced as function number episodes. Crucially, only enhancement these representations, not lower-order predicted power-law Simultaneously, sessions, became stable earlier time aligned, redundant correlated offline gain reducing switch costs. Thus, dynamic task-tailored tesselate space, taming their high-dimensionality.

Language: Английский

Citations

1
Understanding cognitive processes across spatial scales of the brain DOI
Hayoung Song, JeongJun Park, Monica D. Rosenberg

et al.

Trends in Cognitive Sciences, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 1, 2024

Language: Английский

Citations

1
The human posterior parietal cortices orthogonalize the representation of different streams of information concurrently coded in visual working memory DOI Creative Commons
Yaoda Xu

PLoS Biology, Journal Year: 2024, Volume and Issue: 22(11), P. e3002915 - e3002915

Published: Nov. 21, 2024

The key to adaptive visual processing lies in the ability maintain goal-directed representation face of distraction. In working memory (VWM), distraction may come from coding distractors or other concurrently retained targets. This fMRI study reveals a common representational geometry that our brain uses combat both types distractions VWM. Specifically, using pattern decoding, human posterior parietal cortex is shown orthogonalize representations different streams information coded VWM, whether they are targets and distractors, held latter also seen occipitotemporal cortex. Such provides an elegant simple solution enable independent readout, effectively combating information, while accommodating their concurrent representations. scheme differs mechanisms actively suppress block encoding reduce interference. It likely general neural principle supports represent beyond VWM situations where multiple tracked processed simultaneously.

Language: Английский

Citations

1