Accuracy Maximization Analysis for Sensory-Perceptual Tasks: Computational Improvements, Filter Robustness, and Coding Advantages for Scaled Additive Noise

Accuracy Maximization Analysis (AMA) is a recently developed Bayesian ideal observer method for task-specific dimensionality reduction. Given a training set of proximal stimuli (e.g. retinal images), a response noise model, and a cost function, AMA returns the filters (i.e. receptive fields) that extract the most useful stimulus features for estimating a user-specified latent variable from those stimuli. Here, we first contribute two technical advances that significantly reduce AMA’s compute time: we derive gradients of cost functions for which two popular estimators are appropriate, and we implement a stochastic gradient descent (AMA-SGD) routine for filter learning. Next, we show how the method can be used to simultaneously probe the impact on neural encoding of natural stimulus variability, the prior over the latent variable, noise power, and the choice of cost function. Then, we examine the geometry of AMA’s unique combination of properties that distinguish it from better-known statistical methods. Using binocular disparity estimation as a concrete test case, we develop insights that have general implications for understanding neural encoding and decoding in a broad class of fundamental sensory-perceptual tasks connected to the energy model. Specifically, we find that non-orthogonal (partially redundant) filters with scaled additive noise tend to outperform orthogonal filters with constant additive noise; non-orthogonal filters and scaled additive noise can interact to sculpt noise-induced stimulus encoding uncertainty to match task-irrelevant stimulus variability. Thus, we show that some properties of neural response thought to be biophysical nuisances can confer coding advantages to neural systems. Finally, we speculate that, if repurposed for the problem of neural systems identification, AMA may be able to overcome a fundamental limitation of standard subunit model estimation. As natural stimuli become more widely used in the study of psychophysical and neurophysiological performance, we expect that task-specific methods for feature learning like AMA will become increasingly important.Author Summary: In psychophysics and neurophysiology, the stimulus features that are manipulated in experiments are often selected based on intuition, trial-and-error, and historical precedence. Accuracy Maximization Analysis (AMA) is a Bayesian ideal observer method for determining the task-relevant features (i.e. filters) from natural stimuli that nervous systems should select for. In other words, AMA is a method for finding optimal receptive fields for specific tasks. Early results suggest that this method has the potential to be of fundamental importance to neuroscience and perception science. First, we develop AMA-SGD, a new version of AMA that significantly reduces filter-learning time, and use it to learn optimal filters for the classic task of binocular disparity estimation. Then, we find that measureable, task-relevant properties of natural stimuli are the most important determinants of the optimal filters; changes to the prior, cost function, and internal noise have little effect on the filters. Last, we demonstrate that some ubiquitous properties of neural systems, generally thought to be biophysical nuisances, can actually improve the fidelity of neural codes. In particular, we show for the first time that scaled additive noise and redundant (non-orthogonal) filters can interact to sculpt uncertainty due to internal noise to match task-irrelevant natural stimulus variability.