Writing Memories with Light-Addressable Reinforcement Circuitry

Claridge‐Chang, Adam; Roorda, Robert D.; Vrontou, Eleftheria; Sjulson, Lucas; Li, Haiyan; Hirsh, Jay; Miesenböck, Gero

doi:10.1016/j.cell.2009.11.010

Cited by 96 publications

(180 citation statements)

References 0 publications

Supporting

Mentioning

170

Contrasting

Unclassified

Order By: Relevance

“…An SDM operates on high-dimensional binary vectors so that associative and episodic mappings can be learned and retrieved [2,6,26,27,33]. Such high-dimensonal vectors are useful representations of compositional structures [28,29,30,42,45].…”

Section: Introductionmentioning

confidence: 99%

Analogical Mapping with Sparse Distributed Memory: A Simple Model that Learns to Generalize from Examples

Emruli

Sandin

2013

Cogn Comput

View full text Add to dashboard Cite

We present a computational model for the analogical mapping of compositional structures that combines two existing ideas known as holistic mapping vectors and sparse distributed memory. The model enables integration of structural and semantic constraints when learning mappings of the type x i → y i and computing analogies x j → y j for novel inputs x j . The model has a one-shot learning process, is randomly initialized and has three exogenous parameters: the dimensionality D of representations, the memory size S and the probability χ for activation of the memory. After learning three examples the model generalizes correctly to novel examples. We find minima in the probability of generalization error for certain values of χ, S and the number of different mapping examples learned. These results indicate that the optimal size of the memory scales with the number of different mapping examples learned and that the sparseness of the memory is important. The optimal dimensionality of binary representations is of the order 10 4 , which is consistent with a known analytical estimate and the synapse count for most cortical neurons. We demonstrate that the model can learn analogical mappings of generic two-place relationships and we calculate the error probabilities for recall and generalization.

show abstract

Section: Introductionmentioning

confidence: 99%

Analogical Mapping with Sparse Distributed Memory: A Simple Model that Learns to Generalize from Examples

Emruli

Sandin

2013

Cogn Comput

View full text Add to dashboard Cite

show abstract

“…Projection neurons then convey this information to higher order processing centres 4 : the mushroom bodies (MB) and the lateral horn (LH). In contrast, aversive reinforcement signals in response to electric shock are relayed to the MB via dopaminergic neurons [5][6][7] . The olfactory and 3 electric shock signals are integrated in the MB to form aversive olfactory memory 1, 2 .…”

mentioning

confidence: 99%

Mushroom body efferent neurons responsible for aversive olfactory memory retrieval in Drosophila

Séjourné¹,

Plaçais²,

et al. 2011

View full text Add to dashboard Cite

6 These authors contributed equally to this work. 2 Aversive olfactory memory is formed in the Drosophila mushroom bodies (MB).Memory retrieval requires MB output, but it remains unknown how a memory trace in the MB drives conditioned avoidance of a learned odour. To identify neurons involved in olfactory memory retrieval, we performed an anatomical and functional screen of defined sets of MB extrinsic neurons. Here we show that MB-V2 neurons are essential for retrieval of both short-and long-lasting memory, but neither for memory formation nor for memory consolidation. We further show that MB-V2 are cholinergic efferent neurons that project from the MB vertical lobes to the middle superiormedial protocerebrum and the lateral horn (LH). Notably, the odour response of MB-V2 neurons is modified after conditioning. As the LH is implicated in innate responses to repellent odorants, we propose that during memory retrieval, MB-V2 neurons reinforce the olfactory pathway involved in innate odour avoidance.Different odours induce innate approach or avoidance behaviours in Drosophila. Innate odour responses can be modulated by experience, such as associative learning. After simultaneous exposure to an electric shock and an odorant, flies form aversive memory and show robust conditioned odour avoidance that lasts for hours to days, depending on the training protocol [1][2][3] . The neural pathways for odour or shock processing and signal integration in the fly brain have been intensely studied in recent years. Odour information is first represented in the antennal lobes in the form of olfactory receptor neuron activity 4 . Projection neurons then convey this information to higher order processing centres 4 : the mushroom bodies (MB) and the lateral horn (LH). In contrast, aversive reinforcement signals in response to electric shock are relayed to the MB via dopaminergic neurons [5][6][7] . The olfactory and 3 electric shock signals are integrated in the MB to form aversive olfactory memory 1, 2 . The MB are however dispensable for innate avoidance of the repellent odours 8,9 .In adult Drosophila, the MB consist of approximately 2000 Kenyon cells per brain hemisphere, which may be classified into three major types based on their axonal projection: γ neurons, which form only a medial lobe, α/β neurons, whose axons branch to form a vertical (α) and a medial (β) lobe, and α'/β' neurons, which also form a vertical (α') and a medial (β') lobe 10 . Functional brain imaging has revealed localised activation of cAMP/PKA signalling in the MB α lobe in response to simultaneous cholinergic and dopaminergic stimulation 11,12 , that represent respectively the odorant and electric shock pathways.Following associative conditioning, calcium imaging studies have shown that a short-term memory trace is formed in the α'/β' neurons 13, and a long-term one in α lobes 14 . Previous studies have shown that the output of the α/β neurons is necessary for the retrieval of all phases of olfactory memory 15,16 , but the neural circuits that translate ...

show abstract

“…Imaging and behavioral approaches can be combined with light-and heat-activated channels to stimulate or inhibit different neurons in the memory circuit 11,14,16,[22][23][24] to further elucidate their function. Furthermore, mushroom body memory neurons are accessible to whole-cell patch clamp recordings 28 , and mathematical and computational techniques are being used to model Drosophila olfactory memory…”

Section: Discussionmentioning

confidence: 99%

“…The promoter-driven expression of light-or temperature-sensitive transgenes to activate or block the neural activity of specific neurons allows one to investigate which neurons are required for memory acquisition, consolidation and retrieval 3,4,11,15,16,20,[22][23][24] . Memory at 1 hr is typically measured when studying age-related memory impairment because this form of memory appears particularly vulnerable to the effects of ageing [11][12][13] .…”

Section: Introductionmentioning

confidence: 99%