Triplet correlations among similarly tuned cells impact population coding

Area of Science:

• Computational Neuroscience
• Neural Coding
• Systems Neuroscience

Background:

• Neural populations encode stimuli through statistical features of spiking activity.
• Single-cell firing rates and pairwise spike correlations are well-studied coding elements.
• Higher-order spiking correlations (e.g., triplets) are less understood in neural coding.

Purpose of the Study:

• Investigate the impact of higher-order spiking correlations on stimulus encoding.
• Determine if triplet correlations enhance information processing in neural populations.
• Quantify the necessity of including triplet correlations for accurate neural function assessment.

Main Methods:

• Varied triplet correlations in small neural populations (approx. 10 cells) while fixing lower-order statistics.
• Assessed neural population performance on a two-stimulus discrimination task for different triplet correlation levels.
• Utilized geometric analysis to explain findings based on response distribution skew.

Main Results:

• Modulations in triplet correlations significantly enhanced neural coding, particularly when differing between stimuli.
• Triplet correlations are essential for accurate quantification of neural population function when they vary by stimulus.
• Pairwise models offer reasonable accuracy when triplet correlations are similar for both stimuli.

Conclusions:

• Higher-order spiking correlations play a significant role in neural information processing.
• The impact of triplet correlations depends on their stimulus-specific modulation.
• Future experiments require sufficient recording times to accurately measure these complex spiking correlations.