FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3129869075> ?p ?o ?g. }

• W3129869075 endingPage "329" @default.
• W3129869075 startingPage "316" @default.
• W3129869075 abstract "Three core principles of anterior cingulate cortex (ACC) function are reviewed: hierarchy, world models, and cost. Four neural considerations regarding the biophysical implementation of these principles are discussed: modularity, binding, encoding, and learning and regulation. These observations suggest that ACC motivates hierarchical model-based hierarchical reinforcement learning (HMB-HRL). Despite continual debate for the past 30 years about the function of anterior cingulate cortex (ACC), its key contribution to neurocognition remains unknown. However, recent computational modeling work has provided insight into this question. Here we review computational models that illustrate three core principles of ACC function, related to hierarchy, world models, and cost. We also discuss four constraints on the neural implementation of these principles, related to modularity, binding, encoding, and learning and regulation. These observations suggest a role for ACC in hierarchical model-based hierarchical reinforcement learning (HMB-HRL), which instantiates a mechanism motivating the execution of high-level plans. Despite continual debate for the past 30 years about the function of anterior cingulate cortex (ACC), its key contribution to neurocognition remains unknown. However, recent computational modeling work has provided insight into this question. Here we review computational models that illustrate three core principles of ACC function, related to hierarchy, world models, and cost. We also discuss four constraints on the neural implementation of these principles, related to modularity, binding, encoding, and learning and regulation. These observations suggest a role for ACC in hierarchical model-based hierarchical reinforcement learning (HMB-HRL), which instantiates a mechanism motivating the execution of high-level plans. a computer model comprising neuron-like units that are connected via artificial dendrites and axons. Activation is passed via these connections; learning comprises changing of the strengths of the connections. a state that separates regions of state space; for example, a doorway that separates different rooms. An agent must pass through the bottleneck to travel from one region to the other. across-time-averaged reward associated with specific stimuli or actions. Learning cached values is time costly but places minimal burden on the cognitive agent during action execution. RL architecture where a MF system defines the policy, and a MB system trains the MF system. Training typically occurs during periods where the MF system is offline. an option that traverses a principal direction of a learned representation of the environment, which is discovered by following an intrinsic reward function that encourages exploration. Eigenoptions operate at different timescales and can be easily sequenced. HRL abstracts across the action space via an option space, where each option combines a set of (primitive) actions. For example, on a road trip ‘go left’ might be an action while ‘go to Ghent’ might be an option. HMB abstracts across the stimulus or state space in a similar manner. For example, on the same road trip, specific locations in a village may constitute the states and the villages (or even regions of villages) the higher-order states across which HMB plans are made. as used in neuroscience, a semi-transient neural state that persists for a period of time outside the system’s natural equilibrium state. acting and learning to maximize reinforcement, using a model of the world (i.e., a model of the state transitions in the world). acting and learning maximize reinforcement, without a model of the world. This approach uses cached values. as used in cognitive neuroscience, localized neural communities that compute functions that are unique to their group. a system of elements (e.g., neurons) that evolves over time according to nonlinear differential equations. an algorithm that specifies an agent’s behavior for any state of the environment. difference between an event (e.g., reward delivery) and a prediction of that event (e.g., a prediction of reward). an ANN with feedback connections. acting and learning to maximize reinforcement. one of the simplest error-drive learning rules. Its generalization toward multilayer models is the backpropagation algorithm. quantification of the extent to which two or more task-incompatible responses are simultaneously active. short-duration, high-frequency neural oscillations associated with memory retrieval and consolidation. the delay in response times when switching to a new task compared with repeating the same task." @default.
• W3129869075 created "2021-03-01" @default.
• W3129869075 creator A5012480186 @default.
• W3129869075 creator A5054766869 @default.
• W3129869075 date "2021-04-01" @default.
• W3129869075 modified "2023-10-09" @default.
• W3129869075 title "The Best Laid Plans: Computational Principles of Anterior Cingulate Cortex" @default.
• W3129869075 cites W1502005084 @default.
• W3129869075 cites W1617678271 @default.
• W3129869075 cites W1902335101 @default.
• W3129869075 cites W1948186292 @default.
• W3129869075 cites W1973275441 @default.
• W3129869075 cites W1995810302 @default.
• W3129869075 cites W2000998192 @default.
• W3129869075 cites W2009104998 @default.
• W3129869075 cites W2017539895 @default.
• W3129869075 cites W2028384357 @default.
• W3129869075 cites W2028418738 @default.
• W3129869075 cites W2037174944 @default.
• W3129869075 cites W2046194382 @default.
• W3129869075 cites W2048847102 @default.
• W3129869075 cites W2050040379 @default.
• W3129869075 cites W2050963744 @default.
• W3129869075 cites W2057733150 @default.
• W3129869075 cites W2076063813 @default.
• W3129869075 cites W2103660272 @default.
• W3129869075 cites W2127770685 @default.
• W3129869075 cites W2131802924 @default.
• W3129869075 cites W2132130244 @default.
• W3129869075 cites W2133105703 @default.
• W3129869075 cites W2135572147 @default.
• W3129869075 cites W2147689640 @default.
• W3129869075 cites W2149892666 @default.
• W3129869075 cites W2167031142 @default.
• W3129869075 cites W2203267307 @default.
• W3129869075 cites W2281896430 @default.
• W3129869075 cites W2293146851 @default.
• W3129869075 cites W2346736747 @default.
• W3129869075 cites W2400568150 @default.
• W3129869075 cites W2414006382 @default.
• W3129869075 cites W2424347275 @default.
• W3129869075 cites W2461037846 @default.
• W3129869075 cites W2486514532 @default.
• W3129869075 cites W2523721613 @default.
• W3129869075 cites W2523836484 @default.
• W3129869075 cites W2525194628 @default.
• W3129869075 cites W2525444040 @default.
• W3129869075 cites W2553325299 @default.
• W3129869075 cites W2578616969 @default.
• W3129869075 cites W2594142746 @default.
• W3129869075 cites W2594337449 @default.
• W3129869075 cites W2594698283 @default.
• W3129869075 cites W2600886565 @default.
• W3129869075 cites W2609213948 @default.
• W3129869075 cites W2617647752 @default.
• W3129869075 cites W2620631745 @default.
• W3129869075 cites W2736983092 @default.
• W3129869075 cites W2738724892 @default.
• W3129869075 cites W2742829545 @default.
• W3129869075 cites W2744744411 @default.
• W3129869075 cites W2769261924 @default.
• W3129869075 cites W2773963112 @default.
• W3129869075 cites W2785655403 @default.
• W3129869075 cites W2792711916 @default.
• W3129869075 cites W2794448029 @default.
• W3129869075 cites W2800566182 @default.
• W3129869075 cites W2801995417 @default.
• W3129869075 cites W2805663758 @default.
• W3129869075 cites W2854445389 @default.
• W3129869075 cites W2886761031 @default.
• W3129869075 cites W2889901284 @default.
• W3129869075 cites W2896605400 @default.
• W3129869075 cites W2898202513 @default.
• W3129869075 cites W2898268489 @default.
• W3129869075 cites W2902907165 @default.
• W3129869075 cites W2904676870 @default.
• W3129869075 cites W2906699448 @default.
• W3129869075 cites W2908124316 @default.
• W3129869075 cites W2911943505 @default.
• W3129869075 cites W2912387942 @default.
• W3129869075 cites W2921097704 @default.
• W3129869075 cites W2923634585 @default.
• W3129869075 cites W2926786214 @default.
• W3129869075 cites W2932752459 @default.
• W3129869075 cites W2938321354 @default.
• W3129869075 cites W2944341921 @default.
• W3129869075 cites W2946688411 @default.
• W3129869075 cites W2949369413 @default.
• W3129869075 cites W2949641544 @default.
• W3129869075 cites W2950457889 @default.
• W3129869075 cites W2950487657 @default.
• W3129869075 cites W2951065015 @default.
• W3129869075 cites W2954020015 @default.
• W3129869075 cites W2963305465 @default.
• W3129869075 cites W2969548943 @default.
• W3129869075 cites W2971592235 @default.
• W3129869075 cites W2976241900 @default.
• W3129869075 cites W2978368159 @default.

• next