Cortical networks involved in coding and processing of information are constantly shaped by a large variety of plasticity mechanisms. But how do different forms of plasticity interact to shape the structure, dynamics and computational properties of recurrent spiking networks? We use a simple recurrent spiking neural network made of threshold units that allows us to look at the network structure and dynamics in a detailed fashion. Two forms of neuronal plasticity are considered: spike timing dependent plasticity (STDP) that changes synaptic strength and intrinsic plasticity (IP) that changes the excitability of individual neurons to maintain homeostasis of their activity. In analogy to liquid state machines[1] we analyze the ability of such networks to exhibit a fading memory of external inputs and study the extent in which they may discover structure in non-random, predictable time series. We find that STDP and IP interact in non-trivial ways such that the effect of one of them on network behavior can be substantially altered by the presence of the other (also see [2,3]). Specifically, autonomous networks without input shaped by a combination of STDP and IP lead to many limit cycles with stable network behavior in the presence of small perturbations. When we study input driven networks, the causal nature of the STDP rule allows the reservoir to learn structure in the time sequences corresponding to likely sequences of external inputs. The intrinsic plasticity enforces balanced dynamics that utilizes all resources in the network. Together, these mechanisms allow recall (t 0) with a performance that was similar to that of randomly structured reservoirs, while networks trained with just STDP or IP separately perform on average significantly worse (see figure). These differences may be explained by our finding that the two forms of plasticity keep the dynamics of the network at the edge of chaos. Our results underscore the importance of studying the interaction of different forms of plasticity on network behaviour.
We present a non-parametric and computationally efficient method named NeuroXidence that detects coordinated firing of two or more neurons and tests whether the observed level of coordinated firing is significantly different from that expected by chance. The method considers the full auto-structure of the data, including the changes in the rate responses and the history dependencies in the spiking activity. Also, the method accounts for trial-by-trial variability in the dataset, such as the variability of the rate responses and their latencies. NeuroXidence can be applied to short data windows lasting only tens of milliseconds, which enables the tracking of transient neuronal states correlated to information processing. We demonstrate, on both simulated data and single-unit activity recorded in cat visual cortex, that NeuroXidence discriminates reliably between significant and spurious events that occur by chance.
Electroencephalography (EEG) is an essential method used across diverse fields, including neurological diagnosis, cognitive neuroscience, sleep research, and mental health studies. It enables the investigation of neurophysiological functions by recording the brain's electrical activity. A wide variety of EEG and mobile-EEG systems are available on the market. However, adherence to the standards set by the International Federation of Clinical Neurophysiology (IFCN) is essential for ensuring high-quality data collection in clinical environments. The DreamMachine, a mobile EEG device that fully meets these standards, offers 24-channel recordings at a 250 Hz sampling rate, Bluetooth Low Energy (BLE), and additional capabilities to capture electrooculography (EOG) and electrocardiography (ECG) signals. With its low cost, it presents an affordable solution for EEG recording. The software architecture of the open-source DreamMachine is detailed in this study. Focus is placed on data compression and communication between the device and its companion Android application. The details of the Android application's features, including gain settings, bits per channel, filters, bit-shifting, and safety factors are investigated. Subsequently, the system's performance is evaluated through a standard eyes open/closed experiment, comparing its results with a laboratory EEG system across a significant number of participants to assess the performance of the DreamMachine system.
The question of how self-driving cars should behave in dilemma situations has recently attracted a lot of attention in science, media and society. A growing number of publications amass insight into the factors underlying the choices we make in such situations, often using forced-choice paradigms closely linked to the trolley dilemma. The methodology used to address these questions, however, varies widely between studies, ranging from fully immersive virtual reality settings to completely text-based surveys. In this paper we compare virtual reality and text-based assessments, analyzing the effect that different factors in the methodology have on decisions and emotional response of participants. We present two studies, comparing a total of six different conditions varying across three dimensions: The level of abstraction, the use of virtual reality, and time-constraints. Our results show that the moral decisions made in this context are not strongly influenced by the assessment, and the compared methods ultimately appear to measure very similar constructs. Furthermore, we add to the pool of evidence on the underlying factors of moral judgment in traffic dilemmas, both in terms of general preferences, i.e., features of the particular situation and potential victims, as well as in terms of individual differences between participants, such as their age and gender.
Autonomous vehicles as cognitive agents will be an important use case of artificial intelligence in modern societies. Investigating how to increase acceptance and trust, we created a self-explaining car, informing passengers before actions in virtual reality. This study investigates the attitude towards self-driving cars with data from 7850 participants. We show how gender and age affect the attitude towards autonomous vehicles, resulting in female participants being generally less trusting of overall conditions than male participants and a general decrease of acceptance with increasing age. Surprisingly, a self-explaining car providing the passenger with crucial traffic information, although it has a positive impact on trust but influences the intention of using such a car negatively. Therefore, we argue for a highly individualizable in-car communication that meets the adversarial needs of different demographic groups to enable human-machine interactions that foster safe traffic behavior and increase trust and the willingness to use such technology.
Ising models are routinely used to quantify the second order, functional structure of neural populations. With some recent exceptions, they generally do not include the influence of time varying stimulus drive. Yet if the dynamics of network function are to be understood, time varying stimuli must be taken into account. Inclusion of stimulus drive carries a heavy computational burden because the partition function becomes stimulus dependent and must be separately calculated for all unique stimuli observed. This potentially increases computation time by the length of the data set. Here we present an extremely fast, yet simply implemented, method for approximating the stimulus dependent partition function in minutes or seconds. Noting that the most probable spike patterns (which are few) occur in the training data, we sum partition function terms corresponding to those patterns explicitly. We then approximate the sum over the remaining patterns (which are improbable, but many) by casting it in terms of the stimulus modulated missing mass (total stimulus dependent probability of all patterns not observed in the training data). We use a product of conditioned logistic regression models to approximate the stimulus modulated missing mass. This method has complexity of roughly O(LNNpat) where is L the data length, N the number of neurons and N pat the number of unique patterns in the data, contrasting with the O(L2 (N) ) complexity of alternate methods. Using multiple unit recordings from rat hippocampus, macaque DLPFC and cat Area 18 we demonstrate our method requires orders of magnitude less computation time than Monte Carlo methods and can approximate the stimulus driven partition function more accurately than either Monte Carlo methods or deterministic approximations. This advance allows stimuli to be easily included in Ising models making them suitable for studying population based stimulus encoding.
Comparative biology can offer important insights into the evolution of dynamic coordination in the brain. This chapter explores the neural machinery and computations shared by all nervous systems across the animal kingdom, taking into account the fact the complex relationships in coordination between nervous systems and the behavior they produce. It discusses the comparative approaches used to probe brain structure and function, and examines whether there are any fundamental “phase transitions” occurring across groups of organisms in the basic components that build neural circuits and in the kind of computations that these can perform. It also considers the fundamental unity of the functional aspects of neurons, neural circuits, and neural computations in animals such as mammals and birds. Finally, it explains how brain similarities and differences can be used to elucidate complex brain-behavior relationships.
