Modality-specific and modality-independent neural representations work in concert in predictive processes during sequence learning

Cerebral Cortex, Journal Year: 2023, Volume and Issue: 33(12), P. 7783 - 7796

Published: March 21, 2023

Abstract Probabilistic sequence learning supports the development of skills and enables predictive processing. It remains contentious whether visuomotor is driven by representation visual (perceptual coding) or response (motor coding). Neurotypical adults performed a task. Learning occurred incidentally as it was evidenced faster responses to high-probability than low-probability targets. To uncover neurophysiology process, we conducted both univariate analyses multivariate pattern (MVPAs) on temporally decomposed EEG signal. Univariate showed that modulated amplitudes motor code signal but not in perceptual perceptual-motor signals. However, MVPA revealed all 3 codes contribute neurophysiological learnt probabilities. Source localization involvement wider network frontal parietal activations were distinctive across coding levels. These findings suggest sequential regularities rather neither–nor distinction. Moreover, modality-specific encoding worked concert with modality-independent representations, which suggests probabilistic nonunitary encompasses set principles.

Language: Английский

---

Finding Pattern in the Noise: Persistent Implicit Statistical Knowledge Impacts the Processing of Unpredictable Stimuli

Journal of Cognitive Neuroscience, Journal Year: 2024, Volume and Issue: 36(7), P. 1239 - 1264

Published: Jan. 1, 2024

Abstract Humans can extract statistical regularities of the environment to predict upcoming events. Previous research recognized that implicitly acquired knowledge remained persistent and continued influence behavior even when were no longer present in environment. Here, an fMRI experiment, we investigated how persistence is represented brain. Participants (n = 32) completed a visual, four-choice, RT task consisting regularities. Two types blocks constantly alternated with one another throughout task: predictable block type unpredictable ones other. unaware their changing distribution across blocks. Yet, they showed significant at behavioral level not only but also ones, albeit smaller extent. Brain activity range cortical subcortical areas, including early visual cortex, insula, right inferior frontal gyrus, globus pallidus/putamen contributed acquisition The hippocampus as well bilateral angular gyrus seemed play role maintaining this knowledge. results altogether suggest could be exploited relevant, context transmitted retrieved irrelevant without structure.

Language: Английский

---

Lexical surprisal shapes the time course of syntactic structure building

Neurobiology of Language, Journal Year: 2024, Volume and Issue: 5(4), P. 942 - 980

Published: Jan. 1, 2024

When we understand language, recognize words and combine them into sentences. In this article, explore the hypothesis that listeners use probabilistic information about to build syntactic structure. Recent work has shown lexical probability structure both modulate delta-band (<4 Hz) neural signal. Here, investigated whether encoding of changes as a function distributional properties word. To end, analyzed MEG data 24 native speakers Dutch who listened three fairytales with total duration 49 min. Using temporal response functions cumulative model-comparison approach, evaluated contributions features variance in This revealed surprisal values (a feature), well bottom-up node counts feature) positively contributed model Subsequently, compared responses feature between high- low-surprisal values. delay consequence value word: high-surprisal were associated delayed by 150-190 ms. The was not affected word duration, did have origin. These findings suggest brain uses infer structure, highlight an importance for role time process.

Language: Английский

---

Similarity of brain activity patterns during learning and subsequent resting state predicts memory consolidation

Cortex, Journal Year: 2024, Volume and Issue: 179, P. 168 - 190

Published: Aug. 14, 2024

Spontaneous reactivation of brain activity from learning to a subsequent off-line period has been implicated as neural mechanism underlying memory consolidation. However, similarities in may also emerge result individual, trait-like characteristics. Here, we introduced novel approach for analyzing continuous electroencephalography (EEG) data investigate learning-induced changes well characteristics Thirty-one healthy young adults performed task, and their performance was retested after short (∼1 h) delay. Consolidation two distinct types information (serial-order probability) embedded the task were tested reveal functional networks that uniquely predict respective performance. EEG recorded during pre- post-learning rest periods. To associated with consolidation, quantified connectivity between pre-learning (baseline similarity) (post-learning similarity). While comparable patterns these could indicate similarities, baseline similarity changes, possibly spontaneous reactivation. Higher alpha frequency (8.5–9.5 Hz) better consolidation serial-order information, particularly long-range connections across central parietal sites. The probability delta (2.5–3 specifically more local, short-range connections. Furthermore, there substantial overlap associations performance, suggesting robust (trait-like) differences processes.

Language: Английский

---

A neurophysiological perspective on the integration between incidental learning and cognitive control

Communications Biology, Journal Year: 2023, Volume and Issue: 6(1)

Published: March 27, 2023

Abstract Adaptive behaviour requires interaction between neurocognitive systems. Yet, the possibility of concurrent cognitive control and incidental sequence learning remains contentious. We designed an experimental procedure conflict monitoring that follows a pre-defined unknown to participants, in which either statistical or rule-based regularities were manipulated. show participants learnt differences when stimulus was high. Neurophysiological (EEG) analyses confirmed but also specified behavioural results: nature conflict, type learning, stage information processing jointly determine whether support compete with each other. Especially has potential modulate monitoring. Cognitive can engage cooperative fashion adaptation is challenging. Three replication follow-up experiments provide insights into generalizability these results suggest dependent on multifactorial aspects adapting dynamic environment. The study indicates connecting fields advantageous achieve synergistic view adaptive behaviour.

Language: Английский

---

Disentangling Abstraction from Statistical Pattern Matching in Human and Machine Learning

PLoS Computational Biology, Journal Year: 2023, Volume and Issue: 19(8), P. e1011316 - e1011316

Published: Aug. 25, 2023

The ability to acquire abstract knowledge is a hallmark of human intelligence and believed by many be one the core differences between humans neural network models. Agents can endowed with an inductive bias towards abstraction through meta-learning, where they are trained on distribution tasks that share some structure learned applied. However, because networks hard interpret, it difficult tell whether agents have underlying abstraction, or alternatively statistical patterns characteristic abstraction. In this work, we compare performance in meta-reinforcement learning paradigm which generated from rules. We define novel methodology for building “task metamers” closely match statistics but use different generative process, evaluate both metamer tasks. find perform better at than whereas common architectures typically worse matched metamers. This work provides foundation characterizing machine used future developing machines more human-like behavior.

Language: Английский

---

Probing sensitivity to statistical structure in rapid sound sequences using deviant detection tasks

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

Published: April 23, 2024

Abstract Statistical structures and our ability to exploit them are a ubiquitous component of daily life. Yet, we still do not fully understand how track these sophisticated statistics the role they play in sensory processing. Predictive coding frameworks hypothesize that for stimuli can be accurately anticipated based on prior experience, rely more strongly internal model world “surprised” when expectation is unmet. The current study used this phenomenon probe listeners’ sensitivity probabilistic generated using rapid 50 milli-second tone-pip sequences precluded conscious prediction upcoming stimuli. Over three experiments measured response time deviants frequency outside expected range. Predictable were either triplet-based or network-style structure deviant detection contrasted against same set tones but random, unpredictable order. All found structured enhanced relative random sequences. Additionally, Experiment 2 different instantiations community demonstrate level uncertainty modulated saliency. Finally, 3 placed within an established immediately after transition between communities, where perceptual boundary should generate momentary uncertainty. However, manipulation did impact performance. Together results contexts from statistical modulate processing ongoing auditory signal, leading improved detect unexpected stimuli, consistent with predictive framework.

Language: Английский

---

Neural representations of statistical and rule‐based predictions in Gilles de la Tourette syndrome

Human Brain Mapping, Journal Year: 2024, Volume and Issue: 45(8)

Published: June 1, 2024

Gilles de la Tourette syndrome (GTS) is a disorder characterised by motor and vocal tics, which may represent habitual actions as result of enhanced learning associations between stimuli responses (S-R). In this study, we investigated how adults with GTS healthy controls (HC) learn two types regularities in sequence: statistics (non-adjacent probabilities) rules (predefined order). Participants completed visuomotor sequence task while EEG was recorded. To understand the neurophysiological underpinnings these GTS, multivariate pattern analyses on temporally decomposed signal well sLORETA source localisation method were conducted. We found that people showed superior statistical but comparable rule-based compared to HC participants. Adults had different neural representations for both than adults; specifically, maintained regularity longer more overlap them HCs. Moreover, over time scales, distinct fronto-parietal structures contribute groups. propose hyper-learning consequence altered sensitivity encode complex statistics, might lead actions.

Language: Английский

---

Statistical and sequence learning lead to persistent memory in children after a one-year offline period

Scientific Reports, Journal Year: 2021, Volume and Issue: 11(1)

Published: June 14, 2021

Abstract Extraction of environmental patterns underlies human learning throughout the lifespan and plays a crucial role not only in cognitive but also perceptual, motor, social skills. At least two types regularities contribute to acquiring skills: (1) statistical, probability-based regularities, (2) serial order-based regularities. Memory performance and/or over short periods (from minutes weeks) has been widely investigated across lifespan. However, long-term (months or year-long) memory such knowledge received relatively less attention assessed children yet. Here, we aimed test 1-year offline period neurotypical between age 9 15. Participants performed visuomotor four-choice reaction time task designed measure acquisition simultaneously. Short-term consolidation effects were controlled by retesting their after 5-h delay. They then retested on same 1 year later without any practice sessions. successfully acquired both retained period. The successful retention was independent age. Our study demonstrates that representation remains stable long time. These findings offer indirect evidence for developmental invariance model skill consolidation.

Language: Английский

---

Finding pattern in the noise: Persistent implicit statistical knowledge impacts the processing of unpredictable stimuli

Published: Jan. 12, 2022

Humans can extract statistical regularities of the environment to predict upcoming events. Previous research recognized that implicitly acquired knowledge remained persistent and continued influence behavior even when were no longer present in environment. Here, an fMRI experiment, we investigated how persistence is represented brain. Participants (N = 32) completed a visual four-choice reaction time task consisting regularities. Two types blocks constantly alternated with one another throughout task: Predictable block type unpredictable ones other. unaware their changing distribution across blocks. Yet, they showed significant at behavioral level not only predictable but also ones, albeit smaller extent. Brain activity range cortical subcortical areas, including early cortex, insula, right inferior frontal gyrus, globus pallidus/putamen contributed acquisition The hippocampus as well bilateral angular gyrus seemed play role maintaining this knowledge. results altogether suggest could be exploited relevant, context transmitted retrieved irrelevant without structure.

Language: Английский

---