Search for: All records

Advances in neural recording present increasing opportunities to study neural activity in unprecedented detail. Latent variable models (LVMs) are promising tools for analyzing this rich activity across diverse neural systems and behaviors, as LVMs do not depend on known relationships between the activity and external experimental variables. However, progress with LVMs for neuronal population activity is currently impeded by a lack of standardization, resulting in methods being developed and compared in an ad hoc manner. To coordinate these modeling efforts, we introduce a benchmark suite for latent variable modeling of neural population activity. We curate four datasets of neural spiking activity from cognitive, sensory, and motor areas to promote models that apply to the wide variety of activity seen across these areas. We identify unsupervised evaluation as a common framework for evaluating models across datasets, and apply several baselines that demonstrate benchmark diversity. We release this benchmark through EvalAI. 

Most sensorimotor studies investigating the covariation of populations of neurons in primary motor cortex (M1) have considered only a few trained movements made under highly constrained conditions. However, motor behaviors in daily living happen in a far more complex and varied contexts. It is unclear whether M1 neurons would have different population responses in a more naturalistic, unconstrained setting, including requirements to accommodate multiple limbs and body posture, and more extensive proprioceptive inputs. Here, we recorded M1 spiking signals while a monkey performed hand grasp movements in two different contexts: one in the typical constrained lab setting, and the other while moving freely in a large plastic cage. We compared the covariance patterns of the neural activity during movements across the two contexts. We found that the neural covariation patterns accompanying two different hand grasps in the unconstrained context were largely preserved, while they differed across contexts, even for the same type of grasp. We also found that the M1 population activity was confined to context-dependent neural manifolds, but these manifolds were not completely independent, as some dimensions appeared to be shared across the contexts. These results suggest that the coordinated activity of M1 neurons is strongly dependent on behavioral context, in ways that were not entirely anticipated.