Across the nervous system, certain population spiking patterns are observed far more frequently than others. A hypothesis about this structure is that these collective activity patterns function as population codewords–collective modes–carrying information distinct from that of any single cell. We investigate this phenomenon in recordings of ∼150 retinal ganglion cells, the retina’s output. We develop a novel statistical model that decomposes the population response into modes; it predicts the distribution of spiking activity in the ganglion cell population with high accuracy. We found that the modes represent localized features of the visual stimulus that are distinct from the features represented by single neurons. Modes form clusters of activity states that are readily discriminated from one another. When we repeated the same visual stimulus, we found that the same mode was robustly elicited. These results suggest that retinal ganglion cells’ collective signaling is endowed with a form of error-correcting code–a principle that may hold in brain areas beyond retina.

An appealing new principle for neural population codes is that correlations among neurons organize neural activity patterns into a discrete set of clusters, which can each be viewed as a noise-robust population codeword. Previous studies assumed that these codewords corresponded geometrically with local peaks in the probability landscape of neural population responses. Here, we analyze multiple datasets of the responses of ∼150 retinal ganglion cells and show that local probability peaks are absent under broad, non-repeated stimulus ensembles, which are characteristic of natural behavior. However, we find that neural activity still forms noise-robust clusters in this regime, albeit clusters with a different geometry. We start by defining a soft local maximum, which is a local probability maximum when constrained to a fixed spike count. Next, we show that soft local maxima are robustly present, and can moreover be linked across different spike count levels in the probability landscape to form a ridge. We found that these ridges are comprised of combinations of spiking and silence in the neural population such that all of the spiking neurons are members of the same neuronal community, a notion from network theory. We argue that a neuronal community shares many of the properties of Donald Hebb's classic cell assembly, and show that a simple, biologically plausible decoding algorithm can recognize the presence of a specific neuronal community.It is now clear from ample experimental and theoretical evidence that neural circuits throughout the brain encode and transmit information using large populations of neurons [10,19,21,22,25,28,45,53,58,70]. Yet while the manner in which information is represented by single neurons has been intensively studied [12,50,55], the empirical nature of neural population codes is still a topic of active investigation. The advent of new experimental technologies that enable simultaneous recording from hundreds or even thousands of neurons [23,35,40,65] has opened up exciting new possibilities to study this important question. Fundamental to most conceptual approaches is, as with the single neuron case, characterizing the probability distribution over all neuronal responses. The key additional issue for the multi-neuron scenario however is the nature of correlations among neurons, which fundamentally shapes the probability distribution of population activity.So how do correlations affect the code of large neural populations? There are several ideas that have arisen from the past computational neuroscience literature. The oldest is that positive noise correlations can severely limit the encoded information, because they prevent large populations from averaging over the independent noise of neurons [77]. However, this effect can be minimized if the noise correlations are orthogonal in the space of neural activity to the correlations induced by common stimulation [42,63]. In fact, positive correlations with the right structure can even greatly enhance the encoded information [3]. What these ...

Recent advances in experimental techniques have allowed the simultaneous recordings of populations of hundreds of neurons, fostering a debate about the nature of the collective structure of population neural activity. Much of this debate has focused on the empirical findings of a phase transition in the parameter space of maximum entropy models describing the measured neural probability distributions, interpreting this phase transition to indicate a critical tuning of the neural code. Here, we instead focus on the possibility that this is a first-order phase transition which provides evidence that the real neural population is in a ‘structured’, collective state. We show that this collective state is robust to changes in stimulus ensemble and adaptive state. We find that the pattern of pairwise correlations between neurons has a strength that is well within the strongly correlated regime and does not require fine tuning, suggesting that this state is generic for populations of 100+ neurons. We find a clear correspondence between the emergence of a phase transition, and the emergence of attractor-like structure in the inferred energy landscape. A collective state in the neural population, in which neural activity patterns naturally form clusters, provides a consistent interpretation for our results.

No abstract