Hippocampal phase precession from dual input components.

Phase precession is a well known phenomenon in which a hippocampal place cell will fire action potentials at successively earlier phases (relative to the theta-band oscillations recorded in the local field potential) as an animal moves through the cell's receptive field (also known as a place field). We present a model in which CA1 pyramidal cell spiking is driven by dual input components arising from CA3 and EC3. The receptive fields of these two input components overlap but are offset in space from each other such that as the animal moves through the model place field, action potentials are driven first by the CA3 input component and then the EC3 input component. As CA3 synaptic input is known to arrive in CA1 at a later theta phase than EC3 input (Mizuseki et al., 2009; Montgomery et al., 2009), CA1 spiking advances in phase as the model transitions from CA3-driven spiking to EC3-driven spiking. Here spike phase is a function of animal location, placing our results in agreement with many experimental observations characterizing CA1 phase precession (O'Keefe and Recce, 1993; Huxter et al., 2003; Geisler et al., 2007). We predict that experimental manipulations that dramatically enhance or disrupt activity in either of these areas should have a significant effect on phase precession observed in CA1.

A framework for rigorous evaluation of human performance in human and machine learning comparison studies.

Rigorous comparisons of human and machine learning algorithm performance on the same task help to support accurate claims about algorithm success rates and advances understanding of their performance relative to that of human performers. In turn, these comparisons are critical for supporting advances in artificial intelligence. However, the machine learning community has lacked a standardized, consensus framework for performing the evaluations of human performance necessary for comparison. We demonstrate common pitfalls in a designing the human performance evaluation and propose a framework for the evaluation of human performance, illustrating guiding principles for a successful comparison. These principles are first, to design the human evaluation with an understanding of the differences between human and algorithm cognition; second, to match trials between human participants and the algorithm evaluation, and third, to employ best practices for psychology research studies, such as the collection and analysis of supplementary and subjective data and adhering to ethical review protocols. We demonstrate our framework's utility for designing a study to evaluate human performance on a one-shot learning task. Adoption of this common framework may provide a standard approach to evaluate algorithm performance and aid in the reproducibility of comparisons between human and machine learning algorithm performance.

Crossing the Cleft: Communication Challenges Between Neuroscience and Artificial Intelligence.

Historically, neuroscience principles have heavily influenced artificial intelligence (AI), for example the influence of the perceptron model, essentially a simple model of a biological neuron, on artificial neural networks. More recently, notable recent AI advances, for example the growing popularity of reinforcement learning, often appear more aligned with cognitive neuroscience or psychology, focusing on function at a relatively abstract level. At the same time, neuroscience stands poised to enter a new era of large-scale high-resolution data and appears more focused on underlying neural mechanisms or architectures that can, at times, seem rather removed from functional descriptions. While this might seem to foretell a new generation of AI approaches arising from a deeper exploration of neuroscience specifically for AI, the most direct path for achieving this is unclear. Here we discuss cultural differences between the two fields, including divergent priorities that should be considered when leveraging modern-day neuroscience for AI. For example, the two fields feed two very different applications that at times require potentially conflicting perspectives. We highlight small but significant cultural shifts that we feel would greatly facilitate increased synergy between the two fields.

Gain modulation from background synaptic input.

Gain modulation is a prominent feature of neuronal activity recorded in behaving animals, but the mechanism by which it occurs is unknown. By introducing a barrage of excitatory and inhibitory synaptic conductances that mimics conditions encountered in vivo into pyramidal neurons in slices of rat somatosensory cortex, we show that the gain of a neuronal response to excitatory drive can be modulated by varying the level of "background" synaptic input. Simultaneously increasing both excitatory and inhibitory background firing rates in a balanced manner results in a divisive gain modulation of the neuronal response without appreciable signal-independent increases in firing rate or spike-train variability. These results suggest that, within active cortical circuits, the overall level of synaptic input to a neuron acts as a gain control signal that modulates responsiveness to excitatory drive.

Gain modulation of neuronal responses by subtractive and divisive mechanisms of inhibition.

Gain modulation of neuronal responses is widely observed in the cerebral cortex of both anesthetized and behaving animals. Does this multiplicative effect on neuronal tuning curves require underlying multiplicative mechanisms of integration? We compare the effects of a divisive mechanism of inhibition (noisy excitatory and inhibitory synaptic inputs) with the effects of two subtractive mechanisms (shunting conductance and hyperpolarizing current) on the tuning curves of a model cortical neuron. We find that, although the effects of subtractive inhibition can appear nonlinear, they are accompanied by a change in response threshold and are best described as a vertical shift along the response axis. Increasing noisy synaptic activity divisively scales the model responses, reproducing a response-gain control effect. When mutual inhibition between subpopulations of local neurons is included, the model exhibits a gain modulation effect that is better described as input-gain control. We apply these findings to experimental data by examining how noisy synaptic input may underlie divisive surround suppression and attention-driven gain modulation of neuronal responses in the visual system.

Receiver operating characteristic (ROC) analysis for characterizing synaptic efficacy.

The role of background synaptic activity in cortical processing has recently received much attention. How do individual neurons extract information when embedded in a noisy background? When examining the impact of a synaptic input on postsynaptic firing, it is important to distinguish a change in overall firing probability from a true change in neuronal sensitivity to a particular input (synaptic efficacy) that corresponds to a change in detection performance. Here we study the impact of background synaptic input on neuronal sensitivity to individual synaptic inputs using receiver operating characteristic (ROC) analysis. We use the area under the ROC curve as a measure of synaptic efficacy, here defined as the ability of a postsynaptic action potential to identify a particular synaptic input event. An advantage of using ROC analysis to measure synaptic efficacy is that it provides a measure that is independent of postsynaptic firing rate. Furthermore, changes in mean excitation or inhibition, although affecting overall firing probability, do not modulate synaptic efficacy when measured in this way. Changes in overall conductance also affect firing probability but not this form of synaptic efficacy. Input noise, here defined as the variance of the input current, does modulate synaptic efficacy, however. This effect persists when the change in input variance is coupled with a change in conductance (as would result from changing background activity).

Drivers and modulators from push-pull and balanced synaptic input.

In 1998, Sherman and Guillery proposed that there are two types of inputs to cortical neurons; drivers and modulators. These two forms of input are required to explain how, for example, sensory driven responses are controlled and modified by attention and other internally generated gating signals. One might imagine that driver signals are carried by fast ionotropic receptors, whereas modulators correspond to slower metabotropic receptors. Instead, we have proposed a novel mechanism by which both driver and modulator inputs could be carried by transmission through the same types of ionotropic receptors. In this scheme, the distinction between driver and modulator inputs is functional and changeable rather than anatomical and fixed. Driver inputs are carried by excitation and inhibition acting in a push-pull manner. This means that increases in excitation are accompanied by decreases in inhibition and vice versa. Modulators correspond to excitation and inhibition that covary so that they increase or decrease together. Theoretical and experimental work has shown that such an arrangement modulates the gain of a neuron, rather than driving it to respond. Constructing drivers and modulators in this manner allows individual excitatory synaptic inputs to play either role, and indeed to switch between roles, depending on how they are linked with inhibition.