Topic: metaplasticity

The capacities to learn and to store memories are two of the most important functions of the brain, but these abilities are inherently at odds with each other. On one hand, a neuronal network must be flexible enough to quickly store new information, but on the other it must be stable enough to prevent old memories from being constantly overwritten. Metaplasticity in which the timescale of synaptic modifications made by an individual neuron changes could resolve this flexibility-stability tradeoff. We propose that this strategy may be implemented in the olfactory bulb (OB). A striking, characte

Transcranial Magnetic Stimulation (TMS) offers a non-invasive clinical treatment for intractable neuropsychiatric disorders, including Major Depressive Disorder (MDD). Repetitive TMS (rTMS) involves administering a series of patterned magnetic pulses through the skull to modulate cortical brain activity and induce synaptic plasticity in aberrant neural circuitry. When successful, rTMS results in lasting therapeutic changes that may reflect an amelioration in cognitive symptoms. The salient parameters that comprise rTMS protocols, such as pulse frequency, amplitude, etc., interact to determine

Metaplasticity is the change in plasticity of a synapse, and an important mechanism for resolving the stability-plasticity dilemma. However, the exact synaptic or neuron properties that modulate metaplasticity still remains unclear. In this work we use meta-continual learning to discover the importance of factors that contribute to metaplasticity. We use a linear model to assign the relative contribution of four commonly used synaptic properties to metaplasticity: the Hessian, gradient, magnitude of the weights, and activity of the post-synaptic neuron. The coefficients of the linear model ar

Continual learning involves a network adapting to new data while retaining previously learned information. In Artificial Neural Networks , this process faces the challenge of catastrophic forgetting, where updating parameters with new data causes an exponential decline in accuracy over time for previously learned tasks. Several methods have been proposed to address this problem and enable continual learning on future neuromorphic hardware. One such neuroscience-inspired method is metaplasticity, which refers to a network's ability to change the learning capacity of its synapses based on its cu