HHMI Janelia Research Campus, Ashburn, VA, USA.
UCL Gatsby Computational Neuroscience Unit, University College London, London, UK.
Nature. 2019 Jul;571(7765):361-365. doi: 10.1038/s41586-019-1346-5. Epub 2019 Jun 26.
A neuronal population encodes information most efficiently when its stimulus responses are high-dimensional and uncorrelated, and most robustly when they are lower-dimensional and correlated. Here we analysed the dimensionality of the encoding of natural images by large populations of neurons in the visual cortex of awake mice. The evoked population activity was high-dimensional, and correlations obeyed an unexpected power law: the nth principal component variance scaled as 1/n. This scaling was not inherited from the power law spectrum of natural images, because it persisted after stimulus whitening. We proved mathematically that if the variance spectrum was to decay more slowly then the population code could not be smooth, allowing small changes in input to dominate population activity. The theory also predicts larger power-law exponents for lower-dimensional stimulus ensembles, which we validated experimentally. These results suggest that coding smoothness may represent a fundamental constraint that determines correlations in neural population codes.
当神经元群体的刺激反应具有高维度且不相关时，其信息编码效率最高；而当刺激反应具有低维度且相关时，其信息编码稳健性最强。在此，我们分析了在清醒小鼠的视觉皮层中，大量神经元对自然图像的编码的维度。诱发的群体活动具有高维度，且相关性符合一种出人意料的幂律关系：第 n 个主成分方差与 1/n 成比例。这种标度关系并非源自自然图像的幂律谱，因为在刺激白化后它仍然存在。我们从数学上证明，如果方差谱衰减得更慢，那么群体代码就不可能是平滑的，这使得输入中的微小变化能够主导群体活动。该理论还预测，对于低维度的刺激集合，幂律指数会更大，我们通过实验验证了这一点。这些结果表明，编码平滑性可能代表了一种基本约束，它决定了神经群体编码中的相关性。
