Abstract:The Atacama Large Millimeter Array (ALMA) is a joint project between astronomical organizations in Europe, USA and Japan. ALMA will consist of at least 64 12-meter antennas operating in the millimeter and sub-millimeter wavelength range, with baselines up to 10 km. It will be located at an altitude above 5000m in the Chilean Atacama desert[1]. The ALMA Common Software (ACS) provides a software infrastructure common to all partners and consists of a documented collection of common patterns in control systems and of components, which implement those patterns. The heart of ACS is an object model of controlled devices, called Distributed Objects (DOs), implemented as CORBA network objects. Components such as antenna mount, power supply, etc. are defined by means of DOs. A code generator creates Java Bean components for each DO. Programmers can write Java client applications by connecting those Beans with data-manipulation and visualization Beans using commercial visual development tools or programmatically. ACS is based on the experience accumulated with similar projects in the astronomical and particle accelerator contexts, reusing and extending concepts and components. Although designed for ALMA, ACS has the potential for being used in other new control systems and other distributed software projects, since it implements proven design patterns using state of the art, stable and reliable technology.

What problem does this paper attempt to address?

This paper aims to solve the problems of memory formation and input representation in neural networks, especially in oscillatory neural networks such as the hippocampus and olfactory cortex. Specifically, the researchers explored how to use synaptic plasticity (especially spike - time - dependent plasticity) to form memory and input representation in oscillatory neural networks. By introducing a generalized Hebbian learning model, the paper studied how this learning mechanism affects the dynamic characteristics of the network and further explored how these dynamic characteristics in turn affect the computational function of the network. ### Problems the paper attempts to solve 1. **Memory formation and input representation**: - How can spike - time - dependent synaptic plasticity be used to form memory in oscillatory neural networks? - How does the network form memory by clamping external teaching signals during the learning phase and classify or map new inputs during the recall phase? 2. **Constraints of synaptic plasticity**: - How do different forms of synaptic plasticity functions \( A(\tau) \) affect the network's memory formation and input representation capabilities? - How do the experimentally observed synaptic plasticity characteristics (such as LTP and LTD) limit the network's computational function? 3. **Interaction between network dynamics and function**: - How does oscillatory activity affect synaptic changes, and how do these changes in turn affect the network's oscillatory characteristics and its function? - How does a specific form of synaptic plasticity affect the network's resonance frequency and response characteristics? ### Main research contents - **Model introduction**: - Propose an E - I network model containing excitatory and inhibitory neurons, in which synaptic connections are asymmetric. - Describe the oscillatory activity driven by external input and study the form of long - term synaptic changes \( A(\tau) \). - **Learning rules**: - Introduce a spike - time - dependent learning rule that describes the change of synaptic strength over time. - Through linearization analysis and numerical simulation, study the dynamic behavior of the network during the learning process. - **Network behavior**: - Explore the behavioral differences of the network in the learning and recall phases. - Analyze how different types of nonlinearity (Class I and Class II) affect the network's response characteristics. - **Connection between experiment and theory**: - Combine the model with experimental observations to explore how synaptic plasticity affects the network's function. - Propose directions for future research, including further exploration in both experiment and theory. Through these studies, the paper aims to gain a deep understanding of the memory formation and input representation mechanisms in oscillatory neural networks and how these mechanisms are restricted by synaptic plasticity.

Common Software for the ALMA project

Hebbian Imprinting and Retrieval in Oscillatory Neural Networks

LEARNING IN AN OSCILLATORY CORTICAL MODEL

Associative Memory Model of Hippocampus CA3 Using Spike Response Neurons

Oscillation-Driven Spike-Timing Dependent Plasticity Allows Multiple Overlapping Pattern Recognition in Inhibitory Interneuron Networks

How plasticity shapes the formation of neuronal assemblies driven by oscillatory and stochastic inputs

Coherent noise enables probabilistic sequence replay in spiking neuronal networks

Encoding and retrieval in a model of the hippocampal CA1 microcircuit

Magnetic Resonance Force Microscopy Measurement of Entangled Spin States

A model of hippocampal memory based on an adaptive learning rule of synapses

Geometry and dynamics of representations in a precisely balanced memory network related to olfactory cortex

Fluctuation-driven plasticity allows for flexible rewiring of neuronal assemblies

The functional role of oscillatory dynamics of neocortical circuits: a computational perspective

Emergence and maintenance of modularity in neural networks with Hebbian and anti-Hebbian inhibitory STDP

Oscillations, Timing, Plasticity, and Learning in the Cerebellum

Memory-enriched computation and learning in spiking neural networks through Hebbian plasticity

Preexisting hippocampal network dynamics constrain optogenetically induced place fields

Coordination of entorhinal-hippocampal ensemble activity during associative learning

Attractor networks and memory replay of phase coded spike patterns

A Spiking Neurocomputational Model of High-Frequency Oscillatory Brain Responses to Words and Pseudowords

Dreaming neural networks: Forgetting spurious memories and reinforcing pure ones