A Review of Short-Term, Long-Term, and Memory Consolidation Mechanisms in the Hippocampus and Cerebral Cortex

被引：0

作者：

Yang, Zuomin ^{[1
]}

Khairuddin, Anis Salwa Binti Mohd ^{[1
]}

Chuah, Joon Huang ^{[1
,2
]}

Wong, Wei Ru ^{[1
]}

Noman, Hafiz Muhammad Fahad ^{[1
]}

Putri, Tri Wahyu Oktaviana ^{[1
]}

Xu, Xin ^{[3
]}

Haq, Md Minhazul ^{[1
]}

机构：

[1] Univ Malaya, Fac Engn, Dept Elect Engn, Kuala Lumpur 50603, Malaysia

[2] Southern Univ Coll, Fac Engn & Informat Technol, Skudai 81300, Johor, Malaysia

[3] Univ Malaya, Fac Engn, Dept Mech Engn, Kuala Lumpur 50603, Malaysia

来源：

IEEE ACCESS | 2025年 / 13卷

关键词：

Hippocampus; Encoding; Cerebral cortex; Intelligent agents; Biological information theory; Learning (artificial intelligence); Memory management; Visualization; Neurons; Reviews; Artificial intelligence; biological neural systems; continuous learning; intelligent agents; neural networks; long-term memory; TIMING-DEPENDENT PLASTICITY; SPIKE FREQUENCY ADAPTATION; PATTERN SEPARATION; NEURAL-NETWORKS; ELECTROPHYSIOLOGICAL PROPERTIES; CONTEXTUAL MEMORY; THALAMIC CONTROL; EPISODIC MEMORY; DENTATE GYRUS; SLEEP;

D O I：

10.1109/ACCESS.2025.3555539

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Learning and memory have long been a focal point of research across neuroscience and artificial intelligence. A deep understanding of the neural mechanisms underlying learning and memory is necessary to advance artificial intelligence capabilities. Considering this, we review neuroscience principles about learning and memory in the hippocampus and cerebral cortex (neocortex). Specifically, we present the different stages of learning and memory, including short-term memory (STM), long-term memory (LTM), and memory consolidation. Moreover, this work aims to translate neuroscience's mechanisms of learning and memory to intelligent agents, enabling machines with human-like learning, memory, cognitive, and decision-making functions. We discuss a foundational overview of cutting-edge research on continuous learning and memory in artificial intelligence. In addition, we analyze the limitations and research prospects in long-term and short-term memory for continuous learning, providing a solid theoretical groundwork for future advancements.

引用

页码：63248 / 63283

页数：36

共 481 条

[1] Conditional Channel Gated Networks for Task-Aware Continual Learning [J].

Abati, Davide ;

Tomczak, Jakub ;

Blankevoort, Tijmen ;

Calderara, Simone ;

Cucchiara, Rita ;

Bejnordi, Babak Ehteshami .

2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, :3930-3939

[2]

Abbasi Ali, 2022, PR MACH LEARN RES, V199

[3] Prestimulus theta activity predicts correct source memory retrieval [J].

Addante, Richard J. ;

Watrous, Andrew J. ;

Yonelinas, Andrew P. ;

Ekstrom, Arne D. ;

Ranganath, Charan .

PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 2011, 108 (26) :10702-10707

[4]

Adel T, 2020, Arxiv, DOI arXiv:1911.09514

[5]

Ahn H, 2019, ADV NEUR IN, V32

[6] SS-IL: Separated Softmax for Incremental Learning [J].

Ahn, Hongjoon ;

Kwak, Jihwan ;

Lim, Subin ;

Bang, Hyeonsu ;

Kim, Hyojun ;

Moon, Taesup .

2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021), 2021, :824-833

[7]

Aksan E., 2019, arXiv

[8]

Aljundi R, 2019, Arxiv, DOI arXiv:1908.04742

[9]

Aljundi R, 2019, ADV NEUR IN, V32

[10] Memory Aware Synapses: Learning What (not) to Forget [J].

Aljundi, Rahaf ;

Babiloni, Francesca ;

Elhoseiny, Mohamed ;

Rohrbach, Marcus ;

Tuytelaars, Tinne .

COMPUTER VISION - ECCV 2018, PT III, 2018, 11207 :144-161

← 1 2 3 4 5 6 7 8 9 10 →