Learning with reinforcement prediction errors in a model of the Drosophila mushroom body DOI Creative Commons
James Bennett, Andrew Philippides, Thomas Nowotny

et al.

Nature Communications, Journal Year: 2021, Volume and Issue: 12(1)

Published: May 7, 2021

Abstract Effective decision making in a changing environment demands that accurate predictions are learned about outcomes. In Drosophila , such learning is orchestrated part by the mushroom body, where dopamine neurons signal reinforcing stimuli to modulate plasticity presynaptic body output neurons. Building on previous models, which absolute reinforcement, we propose instead reinforcement prediction errors utilising feedback from We formulate rules minimise errors, verify learn simulations, and postulate connectivity explains more physiological observations than an experimentally constrained model. The augmented models reproduce broad range of conditioning blocking experiments, demonstrate absence does not imply error dependent learning. Our results provide five can be tested using established experimental methods.

Language: Английский

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges DOI
Jianshi Tang, Fang Yuan, Xinke Shen

et al.

Advanced Materials, Journal Year: 2019, Volume and Issue: 31(49)

Published: Sept. 24, 2019

As the research on artificial intelligence booms, there is broad interest in brain-inspired computing using novel neuromorphic devices. The potential of various emerging materials and devices for has attracted extensive efforts, leading to a large number publications. Going forward, order better emulate brain's functions, its relevant fundamentals, working mechanisms, resultant behaviors need be re-visited, understood, connected electronics. A systematic overview biological neural systems given, along with their related critical mechanisms. Recent progress reviewed and, more importantly, existing challenges are highlighted hopefully shed light future directions.

Language: Английский

Citations

613
Distinct Dopamine Receptor Pathways Underlie the Temporal Sensitivity of Associative Learning DOI Creative Commons
Annie Handler, Thomas G.W. Graham, Raphael Cohn

et al.

Cell, Journal Year: 2019, Volume and Issue: 178(1), P. 60 - 75.e19

Published: June 1, 2019

Language: Английский

Citations

226
Active Forgetting: Adaptation of Memory by Prefrontal Control DOI Creative Commons
Michael C. Anderson, Justin C. Hulbert

Annual Review of Psychology, Journal Year: 2020, Volume and Issue: 72(1), P. 1 - 36

Published: Sept. 15, 2020

Over the past century, psychologists have discussed whether forgetting might arise from active mechanisms that promote memory loss to achieve various functions, such as minimizing errors, facilitating learning, or regulating one's emotional state. The decade has witnessed a great expansion in knowledge about brain underlying its varying forms. A core discovery concerns role of prefrontal cortex exerting top-down control over mnemonic activity hippocampus and other structures, often via inhibitory control. New findings reveal processes not only induce specific memories but also can suppress operation more broadly, triggering windows anterograde retrograde amnesia healthy people. Recent work extends nonhuman animals, presaging development multilevel mechanistic account spans cognitive, systems, network, even cellular levels. This reveals how organisms adapt their cognitive goals implications for understanding vulnerability psychiatric disorders.

Language: Английский

Citations

206
The neurobiological foundation of memory retrieval DOI
Paul W. Frankland, Sheena A. Josselyn, Stefan Köhler

et al.

Nature Neuroscience, Journal Year: 2019, Volume and Issue: 22(10), P. 1576 - 1585

Published: Sept. 24, 2019

Language: Английский

Citations

196
The Drosophila Mushroom Body: From Architecture to Algorithm in a Learning Circuit DOI
Mehrab N Modi, Yichun Shuai, Glenn Turner

et al.

Annual Review of Neuroscience, Journal Year: 2020, Volume and Issue: 43(1), P. 465 - 484

Published: April 14, 2020

The Drosophila brain contains a relatively simple circuit for forming Pavlovian associations, yet it achieves many operations common across memory systems. Recent advances have established clear framework learning and revealed the following key operations: a) pattern separation, whereby dense combinatorial representations of odors are preprocessed to generate highly specific, nonoverlapping odor patterns used learning; b) convergence, in which sensory information is funneled small set output neurons that guide behavioral actions; c) plasticity, where changing mapping input requires strong reinforcement signal, also modulated by internal state environmental context; d) modularization, consists multiple parallel traces, distinct stability flexibility exist anatomically well-defined modules within network. Cross-module interactions allow higher-order effects past experience influences future learning. Many these parallels with processes formation action selection more complex brains.

Language: Английский

Citations

172
The Neurobiology of Fear Generalization DOI Creative Commons
Arun Asok, Eric R. Kandel, Joseph B. Rayman

et al.

Frontiers in Behavioral Neuroscience, Journal Year: 2019, Volume and Issue: 12

Published: Jan. 15, 2019

The generalization of fear memories is an adaptive neurobiological process that promotes survival in complex and dynamic environments. When confronted with a potential threat, animal must select appropriate defensive response based on previous experiences are not identical, weighing cues contextual information may predict safety or danger. Like other aspects memory, mediated by the coordinated actions prefrontal, hippocampal, amygdalar, thalamic brain areas. In this review, we describe our current understanding behavioral, neural, genetic, biochemical mechanisms involved fear. Fear hallmark many anxiety stress-related disorders, while its emergence, severity, manifestation sex-dependent. Therefore, to improve dialogue between human studies as well accelerate development effective therapeutics, emphasize need examine both sex differences remote timescales rodent models.

Language: Английский

Citations

153
Distilling Causal Effect of Data in Class-Incremental Learning DOI
Xinting Hu, Kaihua Tang, Chunyan Miao

et al.

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Journal Year: 2021, Volume and Issue: unknown

Published: June 1, 2021

We propose a causal framework to explain the catastrophic forgetting in Class-Incremental Learning (CIL) and then derive novel distillation method that is orthogonal existing anti-forgetting techniques, such as data replay feature/label distillation. first 1) place CIL into framework, 2) answer why happens: effect of old lost new training, 3) how techniques mitigate it: they bring back. Based on we distill Colliding Effect between data, which fundamentally equivalent replay, but without any cost storage. Thanks analysis, can further capture Incremental Momentum stream, removing help retain overwhelmed by effect, thus alleviate class testing. Extensive experiments three benchmarks: CIFAR-100, ImageNet-Sub&Full, show proposed improve various state-of-the-art methods large margin (0.72%–9.06%). 1

Language: Английский

Citations

120
Forgetting as a form of adaptive engram cell plasticity DOI
Tomás J. Ryan, Paul W. Frankland

Nature reviews. Neuroscience, Journal Year: 2022, Volume and Issue: 23(3), P. 173 - 186

Published: Jan. 13, 2022

Language: Английский

Citations

120
Engram neurons: Encoding, consolidation, retrieval, and forgetting of memory DOI Creative Commons

Axel Guskjolen,

Mark S. Cembrowski

Molecular Psychiatry, Journal Year: 2023, Volume and Issue: 28(8), P. 3207 - 3219

Published: June 27, 2023

Tremendous strides have been made in our understanding of the neurobiological substrates memory - so-called "engram". Here, we integrate recent progress engram field to illustrate how neurons transform across "lifespan" a from initial encoding, consolidation and retrieval, ultimately forgetting. To do so, first describe cell-intrinsic properties shape emergence at encoding. Second, highlight these encoding preferentially participate synaptic- systems-level memory. Third, changes during guide neural reactivation facilitate recall. Fourth, mechanisms forgetting, can counteract established consolidation, retrieval. Motivated by experimental results four sections, conclude proposing some conceptual extensions traditional view engram, including broadening cell-type participation within engrams stages. In collection, review synthesizes general principles stages, describes future avenues further understand dynamic engram.

Language: Английский

Citations

45
The IBEX Imaging Knowledge-Base: A Community Resource Enabling Adoption and Development of Immunofluoresence Imaging Methods DOI Open Access
Ziv Yaniv, Ifeanyichukwu U. Anidi, Leanne Arakkal

et al.

Published: April 2, 2025

The iterative bleaching extends multiplexity (IBEX) Knowledge-Base is a central portal for researchers adopting IBEX and related 2D 3D immunofluorescence imaging methods. design of the modeled after efforts in open-source software community includes three facets: development platform (GitHub), static website, service data archiving. facilitates practice open science throughout research life cycle by providing validation recommended non-recommended reagents, e.g., primary secondary antibodies. In addition to reporting negative data, empowers method adoption evolution venue sharing protocols, videos, datasets, software, publications. A dedicated discussion forum fosters sense among while addressing questions not covered published manuscripts. Together, scientists from around world are advancing scientific discovery at faster pace, reducing wasted time effort, instilling greater confidence resulting data.

Language: Английский

Citations

2