Single Unit publications

The neonatal hippocampus exhibits regularly recurring waves of synchronized neuronal activity in vitro . Because active sleep (AS), characterized by bursts of phasic motor activity in the form of myoclonic twitching, may provide conditions that are conducive to activity-dependent development of hippocampal circuits, we hypothesized that the waves of synchronous neuronal activity that have been observed in vitro would be associated with AS-related twitching. Using unanesthetized 1- to 12-d-old rats, we report here that the majority of neurons in CA1 and the dentate gyrus (DG) are significantly more active during AS than during either quiet sleep or wakefulness. Neuronal activity typically occurs in phasic bursts, during which most neurons are significantly cross-correlated both within and across the CA1 and DG fields. All AS-active neurons increase their firing rates during periods of myoclonic twitching of the limbs, and a subset of these neurons exhibit a burst of activity immediately after limb twitches, suggesting that the twitches themselves provide sensory feedback to the infant hippocampus, as occurs in the infant spinal cord and neocortex. Finally, the synchronous bursts of neuronal activity are coupled to the emergence of the AS-related hippocampal gamma rhythm during the first postnatal week, as well as the emergence of the AS-related theta rhythm during the second postnatal week. We hypothesize that the phasic motor events of active sleep provide the developing hippocampus with discrete sensory stimulation that contributes to the development and refinement of hippocampal circuits as well as the development of synchronized interactions between hippocampus and neocortex.

Neurons in the auditory midbrain are sensitive to differences in the timing of sounds at the two ears—an important sound localization cue. We used broadband noise stimuli to investigate the interaural-delay sensitivity of low-frequency neurons in two midbrain nuclei: the inferior colliculus (IC) and the dorsal nucleus of the lateral lemniscus. Noise-delay functions showed asymmetries not predicted from a linear dependence on interaural correlation: a stretching along the firing-rate dimension ( rate asymmetry), and a skewing along the interaural-delay dimension ( delay asymmetry). These asymmetries were produced by an envelope-sensitive component to the response that could not entirely be accounted for by monaural or binaural nonlinearities, instead indicating an enhancement of envelope sensitivity at or after the level of the superior olivary complex. In IC, the skew-like asymmetry was consistent with intermediate-type responses produced by the convergence of ipsilateral peak-type inputs and contralateral trough-type inputs. This suggests a stereotyped pattern of input to the IC. In the course of this analysis, we were also able to determine the contribution of time and phase components to neurons' internal delays. These findings have important consequences for the neural representation of interaural timing differences and interaural correlation—cues critical to the perception of acoustic space.

Sensory systems adapt their neural code to changes in the sensory environment, often on multiple time scales. Here, we report a new form of adaptation in a first-order auditory interneuron (AN2) of crickets. We characterize the response of the AN2 neuron to amplitude-modulated sound stimuli and find that adaptation shifts the stimulus-response curves toward higher stimulus intensities, with a time constant of 1.5 s for adaptation and recovery. The spike responses were thus reduced for low-intensity sounds. We then address the question whether adaptation leads to an improvement of the signal's representation and compare the experimental results with the predictions of two competing hypotheses: infomax, which predicts that information conveyed about the entire signal range should be maximized, and selective coding, which predicts that "foreground" signals should be enhanced while "background" signals should be selectively suppressed. We test how adaptation changes the input-response curve when presenting signals with two or three peaks in their amplitude distributions, for which selective coding and infomax predict conflicting changes. By means of Bayesian data analysis, we quantify the shifts of the measured response curves and also find a slight reduction of their slopes. These decreases in slopes are smaller, and the absolute response thresholds are higher than those predicted by infomax. Most remarkably, and in contrast to the infomax principle, adaptation actually reduces the amount of encoded information when considering the whole range of input signals. The response curve changes are also not consistent with the selective coding hypothesis, because the amount of information conveyed about the loudest part of the signal does not increase as predicted but remains nearly constant. Less information is transmitted about signals with lower intensity.

Auditory neurons are likely adapted to process complex stimuli, such as vocalizations, which contain spectrotemporal modulations. However, basic properties of auditory neurons are often derived from tone pips presented in isolation, which lack spectrotemporal modulations. In this context, it is unclear how to deduce the functional role of auditory neurons from their tone pip-derived tuning properties. In this study, spectrotemporal receptive fields (STRFs) were obtained from responses to multi-tone stimulus ensembles differing in their average spectrotemporal density (i.e., number of tone pips per second). STRFs for different stimulus densities were derived from multiple single-unit activity (MUA) and local field potentials (LFPs), simultaneously recorded in primary auditory cortex of cats. Consistent with earlier studies, we found that the spectral bandwidth was narrower for MUA compared with LFPs. Both neural firing rate and LFP amplitude were reduced when the density of the stimulus ensemble increased. Surprisingly, we found that increasing the spectrotemporal sound density revealed with increasing clarity an over-representation of response peaks at frequencies of ∼3, 5, 10, and 20 kHz, in both MUA- and LFP-derived STRFs. Although the decrease in spectral bandwidth and neural activity with increasing stimulus density can likely be accounted for by forward suppression, the mechanisms underlying the over-representation of the octave-spaced response peaks are unclear. Plausibly, the over-representation may be a functional correlate of the periodic pattern of corticocortical connections observed along the tonotopic axis of cat auditory cortex.

Implantable microfabricated microelectrode arrays represent a versatile and powerful tool to record electrophysiological activity across multiple spatial locations in the brain. Spikes and field potentials, however, correspond to only a fraction of the physiological information available at the neural interface. In urethane-anesthetized rats, microfabricated microelectrode arrays were implanted acutely for simultaneous recording of striatal local field potentials, spikes, and electrically evoked dopamine overflow on the same spatiotemporal scale. During these multi-modal recordings we observed (1) that the amperometric method used to detect dopamine did not significantly influence electrophysiological activity, (2) that electrical stimulation in the medial forebrain bundle (MFB) region resulted in electrochemically transduced dopamine transients in the striatum that were spatially heterogeneous within at least 200 microm, and (3) following MFB stimulation, dopamine levels and electrophysiological activity within the striatum exhibited similar temporal profiles. These neural probes are capable of incorporating customized microelectrode geometries and configurations, which may be useful for examining specific spatiotemporal relationships between electrical and chemical signaling in the brain.

To observe neural activity in animals engaged in natural behavior, it is often desirable to minimize or eliminate restraint of the animal. We have developed a simple system for recording from single units in unrestrained cats. An implant with multiple guide tubes and a tiny microdrive is placed inside the recording chamber. An indwelling Pt-Ir microelectrode is advanced incrementally during recording sessions that occur over a period of weeks or months. Electrodes can be easily replaced. We obtain excellent recording stability, and also have been able to sample extensively from a region of cortex or brain stem in a single animal. The essential electronics have been miniaturized and sewn into a light-weight walking jacket, so that we can collect data from a cat who is not connected to any fixed equipment.

Neurons in the inferior colliculus (IC), one of the major integrative centers of the auditory system, process acoustic information converging from almost all nuclei of the auditory brain stem. During this integration, excitatory and inhibitory inputs arrive to auditory neurons at different time delays. Result of this integration determines timing of IC neuron firing. In the mammalian IC, the range of the first spike latencies is very large (5-50 ms). At present, a contribution of excitatory and inhibitory inputs in controlling neurons' firing in the IC is still under debate. In the present study we assess the role of excitation and inhibition in determining first spike response latency in the IC. Postsynaptic responses were recorded to pure tones presented at neuron's characteristic frequency or to downward frequency modulated sweeps in awake bats. There are three main results emerging from the present study: (1) the most common response pattern in the IC is hyperpolarization followed by depolarization followed by hyperpolarization, (2) latencies of depolarizing or hyperpolarizing responses to tonal stimuli are short (3-7 ms) whereas the first spike latencies may vary to a great extent (4-26 ms) from one neuron to another, and (3) high threshold hyperpolarization preceded long latency spikes in IC neurons exhibiting paradoxical latency shift. Our data also show that the onset hyperpolarizing potentials in the IC have very small jitter (<100 micros) across repeated stimulus presentations. The results of this study suggest that inhibition, arriving earlier than excitation, may play a role as a mechanism for delaying the first spike latency in IC neurons.

Background: The mammalian auditory cortex can be subdivided into various fields characterized by neurophysiological and neuroarchitectural properties and by connections with different nuclei of the thalamus. Besides the primary auditory cortex, echolocating bats have cortical fields for the processing of temporal and spectral features of the echolocation pulses. This paper reports on location, neuroarchitecture and basic functional organization of the auditory cortex of the microchiropteran bat Phyllostomus discolor (family: Phyllostomidae). Results: The auditory cortical area of P. discolor is located at parieto-temporal portions of the neocortex. It covers a rostro-caudal range of about 4800 mum and a medio-lateral distance of about 7000 mum on the flattened cortical surface. The auditory cortices of ten adult P. discolor were electrophysiologically mapped in detail. Responses of 849 units (single neurons and neuronal clusters up to three neurons) to pure tone stimulation were recorded extracellularly. Cortical units were characterized and classified depending on their response properties such as best frequency, auditory threshold, first spike latency, response duration, width and shape of the frequency response area and binaural interactions. Based on neurophysiological and neuroanatomical criteria, the auditory cortex of P. discolor could be subdivided into anterior and posterior ventral fields and anterior and posterior dorsal fields. The representation of response properties within the different auditory cortical fields was analyzed in detail. The two ventral fields were distinguished by their tonotopic organization with opposing frequency gradients. The dorsal cortical fields were not tonotopically organized but contained neurons that were responsive to high frequencies only. Conclusion: The auditory cortex of P. discolor resembles the auditory cortex of other phyllostomid bats in size and basic functional organization. The tonotopically organized posterior ventral field might represent the primary auditory cortex and the tonotopically organized anterior ventral field seems to be similar to the anterior auditory field of other mammals. As most energy of the echolocation pulse of P. discolor is contained in the high-frequency range, the non-tonotopically organized high-frequency dorsal region seems to be particularly important for echolocation.

The dominant cue for localization of low-frequency sounds are microsecond differences in the time-of-arrival of sounds at the two ears [interaural time difference (ITD)]. In mammals, ITD sensitivity is established in the medial superior olive (MSO) by coincidence detection of excitatory inputs from both ears. Hence the relative delay of the binaural inputs is crucial for adjusting ITD sensitivity in MSO cells. How these delays are constructed is, however, still unknown. Specifically, the question of whether inhibitory inputs are involved in timing the net excitation in MSO cells, and if so how, is controversial. These inhibitory inputs derive from the nuclei of the trapezoid body, which have physiological and structural specializations for high-fidelity temporal transmission, raising the possibility that well timed inhibition is involved in tuning ITD sensitivity. Here, we present physiological and pharmacological data from in vivo extracellular MSO recordings in anesthetized gerbils. Reversible blockade of synaptic inhibition by iontophoretic application of the glycine antagonist strychnine increased firing rates and significantly shifted ITD sensitivity of MSO neurons. This indicates that glycinergic inhibition plays a major role in tuning the delays of binaural excitation. We also tonically applied glycine, which lowered firing rates but also shifted ITD sensitivity in a way analogous to strychnine. Hence tonic glycine application experimentally decoupled the effect of inhibition from the timing of its inputs. We conclude that, for proper ITD processing, not only is inhibition necessary, but it must also be precisely timed.

Cochlear implant speech processors transmit temporal features of sound as amplitude modulation of constant-rate electrical pulse trains. This study evaluated the central representation of amplitude modulation in the form of phase-locked firing of neurons in the auditory cortex. Anesthetized pigmented guinea pigs were implanted with cochlear electrode arrays. Stimuli were 254 pulse/s (pps) trains of biphasic electrical pulses, sinusoidally modulated with frequencies of 10–64 Hz and modulation depths of −40 to −5 dB re 100% (i.e., 1–56.2% modulation). Single- and multiunit activity was recorded from multi-site silicon-substrate probes. The maximum frequency for significant phase locking (limiting modulation frequency) was ≥60 Hz for 42% of recording sites, whereas phase locking to pulses of unmodulated pulse trains rarely exceeded 30 pps. The strength of phase locking to frequencies ≥40 Hz often varied nonmonotonically with modulation depth, commonly peaking at modulation depths around −15 to −10 dB. Cortical phase locking coded modulation frequency reliably, whereas a putative rate code for frequency was confounded by rate changes with modulation depth. Group delay computed from the slope of mean phase versus modulation frequency tended to increase with decreasing limiting modulation frequency. Neurons in cortical extragranular layers had lower limiting modulation frequencies than did neurons in thalamic afferent layers. Those observations suggest that the low-pass characteristic of cortical phase locking results from intracortical filtering mechanisms. The results show that cortical neurons can phase lock to modulated electrical pulse trains across the range of modulation frequencies and depths presented by cochlear implant speech processors.

Auditory neurons must represent accurately a wide range of sound levels using firing rates that vary over a far narrower range of levels. Recently, we demonstrated that this "dynamic range problem" is lessened by neural adaptation, whereby neurons adjust their input-output functions for sound level according to the prevailing distribution of levels. These adjustments in input-output functions increase the accuracy with which levels around those occurring most commonly are coded by the neural population. Here, we examine how quickly this adaptation occurs. We recorded from single neurons in the auditory midbrain during a stimulus that switched repeatedly between two distributions of sound levels differing in mean level. The high-resolution analysis afforded by this stimulus showed that a prominent component of the adaptation occurs rapidly, with an average time constant across neurons of 160 ms after an increase in mean level, much faster than our previous experiments were able to assess. This time course appears to be independent of both the timescale over which sound levels varied and that over which sound level distributions varied, but is related to neural characteristic frequency. We find that adaptation to an increase in mean level occurs more rapidly than to a decrease. Finally, we observe an additional, slow adaptation in some neurons, which occurs over a timescale of tens of seconds. Our findings provide constraints in the search for mechanisms underlying adaptation to sound level. They also have functional implications for the role of adaptation in the representation of natural sounds.

Loss of temporal processing is characteristic of age-related loss of speech understanding observed in the elderly. Inhibitory glycinergic circuits provide input onto dorsal cochlear nucleus (DCN) projection neurons which likely serve to modulate excitatory responses to time-varying complex acoustic signals. The present study sought to test the hypothesis that age-related loss of inhibition would compromise the ability of output neurons to encode sinusoidally amplitude modulated (SAM) tones. Extracellular recordings were obtained from young and aged FBN rat DCN putative fusiform cells. Stimuli were SAM tones at three modulation depths (100, 50, and 20%) at 30 dB hearing level with the carrier frequency set to the unit's characteristic frequency. Discharge rate and synchrony were calculated to describe SAM responses. There were significant age-related changes in the shape and peak vector strength [best modulation frequency (BMF)] of temporal modulation transfer functions (tMTFs), with no significant age-related changes in rate modulation transfer functions (rMTFs) at BMF. Young neurons exhibited band-pass tMTFs for most SAM conditions while aged fusiform cells exhibited significantly more low-pass or double-peaked tMTFs. There were significant differences in tMTFs between buildup, pauser-buildup, and wide-chopper temporal response types. Young and aged wide-choppers displayed significantly lower vector strength values than the other two temporal DCN response types. Age-related decreases in the number of pauser-buildup response types and increases in wide-chopper types reported previously, could account, in part, for the observed loss of temporal coding of the aged fusiform cell. Age-related changes in SAM coding were similar to changes observed with receptor blockade of glycinergic inhibition onto fusiform cells and consistent with previously observed age-related loss of endogenous glycine levels and changes in normal adult glycine receptor function. DCN changes in SAM coding could, in part, underpin temporal processing deficits observed in the elderly.

We assessed the spatial-tuning properties of units in the cat's anterior auditory field (AAF) and compared them with those observed previously in the primary (A1) and posterior auditory fields (PAF). Multi-channel, silicon-substrate probes were used to record single- and multi-unit activity from the right hemispheres of alpha-chloralose-anesthetized cats. Spatial tuning was assessed using broadband noise bursts that varied in azimuth or elevation. Response latencies were slightly, though significantly, shorter in AAF than A1, and considerably shorter in both of those fields than in PAF. Compared to PAF, spike counts and latencies were more poorly modulated by changes in stimulus location in AAF and A1, particularly at higher sound pressure levels. Moreover, units in AAF and A1 demonstrated poorer level tolerance than units in PAF with spike rates modulated as much by changes in stimulus intensity as changes in stimulus location. Finally, spike-pattern-recognition analyses indicated that units in AAF transmitted less spatial information, on average, than did units in PAF-an observation consistent with recent evidence that PAF is necessary for sound-localization behavior, whereas AAF is not.

We investigated encoding properties of identified auditory interneurons in two not closely related grasshopper species (Acrididae). The neurons can be homologized on the basis of their similar morphologies and physiologies. As test stimuli, we used the species-specific stridulation signals of Chorthippus biguttulus , which evidently are not relevant for the other species, Locusta migratoria . We recorded spike trains produced in response to these signals from several neuron types at the first levels of the auditory pathway in both species. Using a spike train metric to quantify differences between neuronal responses, we found a high similarity in the responses of homologous neurons: interspecific differences between the responses of homologous neurons in the two species were not significantly larger than intraspecific differences (between several specimens of a neuron in one species). These results suggest that the elements of the thoracic auditory pathway have been strongly conserved during the evolutionary divergence of these species. According to the ‘efficient coding’ hypothesis, an adaptation of the thoracic auditory pathway to the specific needs of acoustic communication could be expected. We conclude that there must have been stabilizing selective forces at work that conserved coding characteristics and prevented such an adaptation.

The lateral superior olive (LSO) is one of the earliest sites in the auditory pathway that is involved in processing acoustical cues to sound location. Here, we tested the hypothesis that LSO neurons can signal small changes in interaural level differences (ILDs), a cue to horizontal sound location, of pure tones based on discharge rate consistent with psychophysical performance in the discrimination of ILDs. Neural thresholds for ILD discrimination were determined from the discharge rates and associated response variability of single units in response to 300 ms tones in the LSO of barbiturate-anesthetized cats using detection theory. Neural response variability was well described by a power function of the mean rate, both in individual neurons and collectively; LSO neurons were less variable than expected from a Poisson process. Compared with psychophysical data, the best-threshold ILDs of single LSO neurons were comparable with or better than behavior over the full range of frequencies (0.3–35 kHz) and pedestal ILDs (±25 dB) explored in this study. With a pedestal ILD of 0 dB, ILD increments of 1 dB could be discriminated by some neurons, with a median of 4.35 dB across neurons. For pedestal ILDs away from 0 dB, the best-threshold ILDs were as low as 0.5 dB, with a median of 2.3 dB. These findings support the hypothesis that the LSO plays a role in the extraction of ILD, and that the representation of ILD by LSO neurons may set a lower bound on the behavioral sensitivity to ILDs.

The primate auditory cortex contains three interconnected regions (core, belt, parabelt), which are further subdivided into discrete areas. The caudomedial area (CM) is one of about seven areas in the belt region that has been the subject of recent anatomical and physiological studies conducted to define the functional organization of auditory cortex. The main goal of the present study was to examine temporal coding in area CM of marmoset monkeys using two related classes of acoustic stimuli: (1) marmoset twitter calls; and (2) frequency-modulated (FM) sweep trains modeled after the twitter call. The FM sweep trains were presented at repetition rates between 1 and 24 Hz, overlapping the natural phrase frequency of the twitter call (6-8 Hz). Multiunit recordings in CM revealed robust phase-locked responses to twitter calls and FM sweep trains. For the latter, phase-locking quantified by vector strength (VS) was best at repetition rates between 2 and 8 Hz, with a mean of about 5 Hz. Temporal response patterns were not strictly phase-locked, but exhibited dynamic features that varied with the repetition rate. To examine these properties, classification of the repetition rate from the temporal response pattern evoked by twitter calls and FM sweep trains was examined by Fisher's linear discrimination analysis (LDA). Response classification by LDA revealed that information was encoded not only by phase-locking, but also other components of the temporal response pattern. For FM sweep trains, classification was best for repetition rates from 2 to 8 Hz. Thus, the majority of neurons in CM can accurately encode the envelopes of temporally complex stimuli over the behaviorally-relevant range of the twitter call. This suggests that CM could be engaged in processing that requires relatively precise temporal envelope discrimination, and supports the hypothesis that CM is positioned at an early stage of processing in the auditory cortex of primates.

Audible frequencies of sound are encoded in a continuous manner along the length of the cochlea, and frequency is transmitted to the brain as a representation of place on the basilar membrane. The resulting tonotopic map has been assumed to be a continuous smooth progression from low to high frequency throughout the central auditory system. Here, physiological and anatomical data show that best frequency is represented in a discontinuous manner in the inferior colliculus, the major auditory structure of the midbrain. Multiunit maps demonstrate a distinct stepwise organization in the order of best frequency progression. Furthermore, independent data from single neurons show that best frequencies at octave intervals of approximately one-third are more prevalent than others. These data suggest that, in the inferior colliculus, there is a defined space of tissue devoted to a given frequency, and input within this frequency band may be pooled for higher-level processing.

In this study, we investigate the ability of the mammalian auditory pathway to adapt its strategy for temporal processing under natural stimulus conditions. We derive temporal receptive fields from the responses of neurons in the inferior colliculus to vocalization stimuli with and without additional ambient noise. We find that the onset of ambient noise evokes a change in receptive field dynamics that corresponds to a change from bandpass to lowpass temporal filtering. We show that these changes occur within a few hundred milliseconds of the onset of the noise and are evident across a range of overall stimulus intensities. Using a simple model, we illustrate how these changes in temporal processing exploit differences in the statistical properties of vocalizations and ambient noises to increase the information in the neural response in a manner consistent with the principles of efficient coding.

Neurons in the superior paraolivary nucleus (SPON) of the rat respond to the offset of pure tones with a brief burst of spikes. Medial nucleus of the trapezoid body (MNTB) neurons, which inhibit the SPON, produce a sustained pure tone response followed by an offset response characterized by a period of suppressed spontaneous activity. This MNTB offset response is duration dependent and critical to the formation of SPON offset spikes [Kadner A, Kulesza RJ Jr, Berrebi AS (2006) Neurons in the medial nucleus of the trapezoid body and superior paraolivary nucleus of the rat may play a role in sound duration coding. J Neurophysiol. 95:1499-1508; Kulesza RJ Jr, Kadner A, Berrebi AS (2007) Distinct roles for glycine and GABA in shaping the response properties of neurons in the superior paraolivary nucleus of the rat. J Neurophysiol 97:1610-1620]. Here we examine the temporal resolution of the rat's MNTB/SPON circuit by assessing its capability to i) detect gaps in tones, and ii) synchronize to sinusoidally amplitude modulated (SAM) tones. Gap detection was tested by presenting two identical pure tone markers interrupted by gaps ranging from 0 to 25 ms duration. SPON neurons responded to the offset of the leading marker even when the two markers were separated only by their ramps (i.e. a 0 ms gap); longer gap durations elicited progressively larger responses. MNTB neurons produced an offset response at gap durations of 2 ms or longer, with a subset of neurons responding to 0 ms gaps. SAM tone stimuli used the unit's characteristic frequency as a carrier, and modulation rates ranged from 40 to 1160 Hz. MNTB neurons synchronized to modulation rates up to approximately 1 kHz, whereas spiking of SPON neurons decreased sharply at modulation rates >or=400 Hz. Modulation transfer functions based on spike count were all-pass for MNTB neurons and low-pass for SPON neurons; the modulation transfer functions based on vector strength were low-pass for both nuclei, with a steeper cutoff for SPON neurons. Thus, the MNTB/SPON circuit encodes episodes of low stimulus energy, such as gaps in pure tones and troughs in amplitude modulated tones. The output of this circuit consists of brief SPON spiking episodes; their potential effects on the auditory midbrain and forebrain are discussed.

How do neuronal populations in the auditory cortex represent acoustic stimuli? Although sound-evoked neural responses in the anesthetized auditory cortex are mainly transient, recent experiments in the unanesthetized preparation have emphasized subpopulations with other response properties. To quantify the relative contributions of these different subpopulations in the awake preparation, we have estimated the representation of sounds across the neuronal population using a representative ensemble of stimuli. We used cell-attached recording with a glass electrode, a method for which single-unit isolation does not depend on neuronal activity, to quantify the fraction of neurons engaged by acoustic stimuli (tones, frequency modulated sweeps, white-noise bursts, and natural stimuli) in the primary auditory cortex of awake head-fixed rats. We find that the population response is sparse, with stimuli typically eliciting high firing rates (>20 spikes/second) in less than 5% of neurons at any instant. Some neurons had very low spontaneous firing rates (<0.01 spikes/second). At the other extreme, some neurons had driven rates in excess of 50 spikes/second. Interestingly, the overall population response was well described by a lognormal distribution, rather than the exponential distribution that is often reported. Our results represent, to our knowledge, the first quantitative evidence for sparse representations of sounds in the unanesthetized auditory cortex. Our results are compatible with a model in which most neurons are silent much of the time, and in which representations are composed of small dynamic subsets of highly active neurons.