Choice-selective sequences dominate in cortical relative to thalamic inputs to nucleus accumbens, providing a potential substrate for credit assignment DOI Open Access

Nathan F. Parker,

Avinash Baidya,

Julia Cox

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2019, Номер unknown

Опубликована: Авг. 6, 2019

How are actions linked with subsequent outcomes to guide choices? The nucleus accumbens, which is implicated in this process, receives glutamatergic inputs from the prelimbic cortex and midline regions of thalamus. However, little known about whether how representations differ across these input pathways. By comparing during a reinforcement learning task mice, we discovered that cortical preferentially represent choices, whereas thalamic cues. Choice-selective activity organized sequences persist beyond outcome. Through computational modeling, demonstrate can support neural implementation algorithms, both circuit model based on synaptic plasticity, one dynamics. Finally, test confirm predictions our models by direct manipulation accumbens neurons. Thus, integrate experiment modeling suggest solutions for credit assignment.

Язык: Английский

A collicular map for touch-guided tongue control DOI
Brendan S. Ito, Yongjie Gao, Brian Kardon

и другие.

Nature, Год журнала: 2025, Номер unknown

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

5

A neural network for information seeking DOI Creative Commons

Jesse White,

Ethan S. Bromberg-Martin, Sarah R. Heilbronner

и другие.

Nature Communications, Год журнала: 2019, Номер 10(1)

Опубликована: Ноя. 14, 2019

Abstract Humans and other animals often show a strong desire to know the uncertain rewards their future has in store, even when they cannot use this information influence outcome. However, it is unknown how brain predicts opportunities gain motivates information-seeking behavior. Here we that neurons network of interconnected subregions primate anterior cingulate cortex basal ganglia predict moment gaining about rewards. Spontaneous increases prediction signals are followed by gaze shifts toward objects associated with resolving uncertainty, pharmacologically disrupting reduces motivation seek information. These findings demonstrate cortico-basal mechanism responsible for motivating actions resolve uncertainty seeking knowledge future.

Язык: Английский

Процитировано

119

A quantitative reward prediction error signal in the ventral pallidum DOI
David J. Ottenheimer, Bilal A. Bari, Elissa Sutlief

и другие.

Nature Neuroscience, Год журнала: 2020, Номер 23(10), С. 1267 - 1276

Опубликована: Авг. 10, 2020

Язык: Английский

Процитировано

89

The multi-dimensional nature of vocal learning DOI Creative Commons
Sonja C. Vernes, Buddhamas Kriengwatana, Veronika C. Beeck

и другие.

Philosophical Transactions of the Royal Society B Biological Sciences, Год журнала: 2021, Номер 376(1836)

Опубликована: Сен. 6, 2021

How learning affects vocalizations is a key question in the study of animal communication and human language. Parallel efforts birds humans have taught us much about how vocal works on behavioural neurobiological level. Subsequent revealed variety cases among mammals which experience also has major influence repertoires. Janik Slater ( Anim. Behav. 60 , 1–11. doi:10.1006/anbe.2000.1410 )) introduced distinction between usage production learning, providing general framework to categorize different types vocalizations. This idea was built by Petkov Jarvis Front. Evol. Neurosci. 4 12. doi:10.3389/fnevo.2012.00012 emphasize more continuous distribution limited complex learners. Yet, with studies empirical data, limits initial frameworks become apparent. We build these refine categorization light advances made since their publication widespread agreement that not binary trait. propose novel classification system, based definitions Slater, deconstructs into dimensions aid understanding mechanisms involved this behaviour. consider can change without considers context specificity timing. identify including copying auditory models (convergence/divergence model sounds, accuracy copying), degree (type breadth learning) timing (when takes place, length time it long retained). grey areas current mechanistic behaviours. Our identifies research needs will help inform evolutionary endeavouring uncover multi-dimensional nature learning. article part theme issue ‘Vocal animals humans’.

Язык: Английский

Процитировано

64

Choice-selective sequences dominate in cortical relative to thalamic inputs to NAc to support reinforcement learning DOI Creative Commons

Nathan F. Parker,

Avinash Baidya,

Julia Cox

и другие.

Cell Reports, Год журнала: 2022, Номер 39(7), С. 110756 - 110756

Опубликована: Май 1, 2022

How are actions linked with subsequent outcomes to guide choices? The nucleus accumbens, which is implicated in this process, receives glutamatergic inputs from the prelimbic cortex and midline regions of thalamus. However, little known about whether how representations differ across these input pathways. By comparing during a reinforcement learning task mice, we discovered that cortical preferentially represent choices, whereas thalamic cues. Choice-selective activity organized sequences persist beyond outcome. Through computational modeling, demonstrate can support neural implementation reinforcement-learning algorithms, both circuit model based on synaptic plasticity one dynamics. Finally, test confirm prediction our models by direct manipulation accumbens neurons.

Язык: Английский

Процитировано

43

Modeling and dissociation of intrinsic and input-driven neural population dynamics underlying behavior DOI
Parsa Vahidi, Omid G. Sani, Maryam M. Shanechi

и другие.

Proceedings of the National Academy of Sciences, Год журнала: 2024, Номер 121(7)

Опубликована: Фев. 9, 2024

Neural dynamics can reflect intrinsic or dynamic inputs, such as sensory inputs from other brain regions. To avoid misinterpreting temporally structured dynamics, dynamical models of neural activity should account for measured inputs. However, incorporating remains elusive in joint modeling neural-behavioral data, which is important studying computations behavior. We first show how training while considering behavior but not input may lead to misinterpretations. then develop an analytical learning method linear that simultaneously accounts activity, behavior, and The provides the capability prioritize behaviorally relevant dissociate them both dynamics. In data a simulated with fixed performs different tasks, correctly finds same regardless task methods be influenced by task. datasets three subjects performing two motor tasks instruction reveals low-dimensional are missed more predictive and/or activity. also uniquely largely similar across whereas overall not. These input-driven uncover otherwise missed.

Язык: Английский

Процитировано

15

Dopaminergic error signals retune to social feedback during courtship DOI
Andrea Roeser, Vikram Gadagkar, Anindita Das

и другие.

Nature, Год журнала: 2023, Номер 623(7986), С. 375 - 380

Опубликована: Сен. 27, 2023

Язык: Английский

Процитировано

23

Discrete Evaluative and Premotor Circuits Enable Vocal Learning in Songbirds DOI Creative Commons
Matthew Gene Kearney, Timothy L. Warren, Erin Hisey

и другие.

Neuron, Год журнала: 2019, Номер 104(3), С. 559 - 575.e6

Опубликована: Авг. 22, 2019

Язык: Английский

Процитировано

52

The neurobiology of innate and learned vocalizations in rodents and songbirds DOI
Richard Mooney

Current Opinion in Neurobiology, Год журнала: 2020, Номер 64, С. 24 - 31

Опубликована: Фев. 18, 2020

Язык: Английский

Процитировано

41

Expression of FoxP2 in the basal ganglia regulates vocal motor sequences in the adult songbird DOI Creative Commons
Lei Xiao, Devin P. Merullo, Therese MI Koch

и другие.

Nature Communications, Год журнала: 2021, Номер 12(1)

Опубликована: Май 11, 2021

Abstract Disruption of the transcription factor FoxP2, which is enriched in basal ganglia, impairs vocal development humans and songbirds. The ganglia are important for selection sequencing motor actions, but circuit mechanisms governing accurate learned vocalizations unknown. Here, we show that expression FoxP2 vital fluent initiation termination birdsong, as well maintenance song syllable adulthood. Knockdown imbalances dopamine receptor across striatal direct-like indirect-like pathways, suggesting a role dopaminergic signaling regulating sequencing. Confirming this prediction, phasic activation, not inhibition, during singing drives repetition syllables, thus also impairing birdsong. These findings demonstrate discrete origins dysfluent elements songbirds, with implications speech disorders.

Язык: Английский

Процитировано

37