Computational roles of plastic probabilistic synapses

Montero, Milton Llera; Sacramento, João; Costa, Rui Ponte

doi:10.1016/j.conb.2018.09.002

Cited by 24 publications

(29 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[263][264][265] Inference is a terminology in Bayes Adv. the firing time of spikes can be modeled as a Poisson process and release of neurotransmitters from synapses are stochastic.…”

Section: Stochastic Synaptic Release and Synaptic Samplingmentioning

confidence: 99%

See 1 more Smart Citation

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges

et al. 2019

View full text Add to dashboard Cite

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

show abstract

“…[263][264][265] Inference is a terminology in Bayes Adv. the firing time of spikes can be modeled as a Poisson process and release of neurotransmitters from synapses are stochastic.…”

Section: Stochastic Synaptic Release and Synaptic Samplingmentioning

confidence: 99%

“…A recent theory has linked this feature to a technique called sampling in Bayesian learning. [263][264][265] Inference is a terminology in Bayes Adv. Mater.…”

Section: Stochastic Synaptic Release and Synaptic Samplingmentioning

confidence: 99%

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges

et al. 2019

View full text Add to dashboard Cite

show abstract

“…The functional role of probabilistic synapses is highly debated [15]. The proposed functional roles include regularization and improved generalization in deep neural networks [96,97] and energy saving constraints [98].…”

Section: Functional Role Of Probabilistic Synapsesmentioning

confidence: 99%

“…We assume a simple synaptic model consisting of a plastic synaptic probability p and a synaptic weight w, which we fix to 1 in order to concentrate on our main ideas. The synaptic probability corresponds to the pre-synaptic neuro-transmitter release probability and the weight w to the post-synaptic quantal amplitude [15]. The neuro-transmitter release probability of synapses in the brain is highly variable and typically between 0.1 and 0.9 [16].…”

Section: Introductionmentioning

confidence: 99%

Adaptive Tuning Curve Widths Improve Sample Efficient Learning

Meier

Dang-Nhu

Steger

2019

Preprint

View full text Add to dashboard Cite

Natural brains perform miraculously well in learning new tasks from a small number of samples, whereas sample efficient learning is still a major open problem in the field of machine learning. Here, we raise the question, how the neural coding scheme affects sample efficiency, and make first progress on this question by proposing and analyzing a learning algorithm that uses a simple reinforce-type plasticity mechanism and does not require any gradients to learn low dimensional mappings. It harnesses three bio-plausible mechanisms, namely, population codes with bell shaped tuning curves, continous attractor mechanisms and probabilistic synapses, to achieve sample efficient learning. We show both theoretically and by simulations that population codes with broadly tuned neurons lead to high sample efficiency, whereas codes with sharply tuned neurons account for high final precision. Moreover, a dynamic adaptation of the tuning width during learning gives rise to both, high sample efficiency and high final precision. We prove a sample efficiency guarantee for our algorithm that lies within a logarithmic factor from the information theoretical optimum. Our simulations show that for low dimensional mappings, our learning algorithm achieves comparable sample efficiency to multi-layer perceptrons trained by gradient descent, although it does not use any gradients. Furthermore, it achieves competitive sample efficiency in low dimensional reinforcement learning tasks. From a machine learning perspective, these findings may inspire novel approaches to improve sample efficiency. From a neuroscience perspective, these findings suggest sample efficiency as a yet unstudied functional role of adaptive tuning curve width.K eywords sample efficiency · neural tuning curves · population codes · gradient-free learning · reinforcement learning 6 if L(x, y) ≥L then c ij = c ij − 1 for all (i, j) ∈ I × J; 7 else c ij = c ij − 1 for all (i, j) ∈ I × J; 8 for (i, j) ∈ I × J do 9 if c ij ≤ 0 then p ij = 0;

show abstract

“…In the cortex, neurons continuously generate irregular spike trains with fluctuating membrane potential and largely varying firing rate even across trials where an animal solves the same task and shows identical responses [1][2][3][4][5][6][7]. Besides neurons, synapses are also stochastic in cerebral cortical [8]. Formation, elimination and volume change of dendritic spines exhibit random fluctuation [9][10][11][12][13][14][15] and transmitter release from synapses is also inherently unreliable [16][17][18].…”

Section: Introductionmentioning

confidence: 99%

Dual stochasticity in the cortex as a biologically plausible learning with the most efficient coding

Teramae

2019

Preprint

View full text Add to dashboard Cite

Neurons and synapses in the cerebral cortex behave stochastically even during precise perception and reliable learning of animals. While studies have revealed advantages of these stochastic features, relationship and synergy of them remain elusive. Here, we show that these stochastic features are unified into a framework to provide an efficient and biologically-plausible learning algorithm that consistently explains various experimental findings of the brain, which includes statistics of cortical circuit and the most efficient power-low encoding of cortex. The algorithm can be regarded as an integration of major branches of existing algorithms of neural networks, back propagation learning and Boltzmann machine, while it overcomes known limitations of them. As an experimentally testable prediction, slow retrograde modulation of excitability of neurons is also derived. We anticipate that the algorithm is, due to its simplicity and flexibility, beneficial to develop neuromorphic devices or scalable AI-chips, and bridges the gap between neuroscience and machine learning.

show abstract

Computational roles of plastic probabilistic synapses

Cited by 24 publications

References 61 publications

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges

Bridging Biological and Artificial Neural Networks with Emerging Neuromorphic Devices: Fundamentals, Progress, and Challenges

Adaptive Tuning Curve Widths Improve Sample Efficient Learning

Dual stochasticity in the cortex as a biologically plausible learning with the most efficient coding

Contact Info

Product

Resources

About