Predictive Coding Explains Asymmetric Connectivity in the Brain: A Neural Network Study

Abstract

Seminal frameworks of predictive coding propose a hierarchy of generative modules, each attempting to infer the neural representation of the module one level below; the predictions are carried by top-down feedback projections, while the predictive error is propagated by reciprocal forward pathways. Such symmetric feedback connections support visual processing of noisy stimuli in computational models. However, neurophysiological studies have yielded evidence of asymmetric cortical feedback connections. We investigated the contribution of neural feedback during sensorimotor processes, in particular visual processing during grasp planning, by utilizing convolutional neural network models that had been augmented with predictive feedback and were trained to compute grasp positions for real-world objects. After establishing an ameliorative effect of symmetric feedback on grasp detection performance when evaluated on noisy stimuli, we characterized the performance effects of asymmetric feedback, similar to that observed in the cortex. Specifically, we tested model variants extended with short-, medium- and long-range feedback connections (i) originating at the same source layer or (ii) terminating at the same target layer. We found that the performance-enhancing effect of predictive coding under adverse conditions was optimal for medium-range asymmetric feedback. Moreover, this effect was most prominent when medium-range feedback originated at a level of representational abstraction that was proximal to the input layer, in contrast to more distal layers. To conclude, our simulations show that introducing biologically realistic asymmetric predictive feedback improves model robustness to noisy visual stimuli in a neural network model optimized for grasp detection. Significance statementIt is commonly held that the brain predicts the causes of its sensorium via top-down neural pathways. While canonical models of predictive coding assume reciprocal feedforward and feedback connections, functional evidence highlights the importance of non-reciprocal asymmetric feedback, whose role remains poorly understood, particularly in sensorimotor functions. Using neural network models of grasp planning, we characterized optimal pathlengths and source regions for asymmetric feedback facilitating visuomotor processing of noisy sensory inputs. Our findings show that medium-range feedback from early layers marks a sweet spot, incorporating optimal distance between the neural representations of source/target layers and representational abstraction of the feedback source. This intimates an uncharted role of intermediate brain areas along the visuomotor stream as a source of predictive signals.