Tirin Moore | Searcharxiv

arXiv2026

STSBench: A Large-Scale Dataset for Modeling Neuronal Activity in the Dorsal Stream of Primate Visual Cortex

The primate visual system is typically divided into two streams - the ventral stream, responsible for object recognition, and the dorsal stream, responsible for encoding spatial relations and motion. Recent studies have shown that convolutional neural networks (CNNs) pretrained on object recognition tasks are remarkably effective at predicting neuronal responses in the ventral stream, shedding light on the neural mechanisms underlying object recognition. However, similar models of the dorsal stream remain underdeveloped due to the lack of large scale datasets encompassing dorsal stream areas. To address this gap, we present STSBench, a dataset of large-scale, single neuron recordings from over 2,000 neurons in the superior temporal sulcus (STS), a nearly 50-fold increase over existing dorsal stream datasets, collected while Rhesus macaques viewed thousands of unique, natural videos. We show that our dataset can be used for benchmarking encoding models of dorsal stream neuronal responses and reconstructing visual input from neural activity.

q-bio.NC↗

arXiv2013

A mathematical model for distinguishing bias from sensitivity effects in multialternative detection tasks

Studies investigating the neural bases of cognitive phenomena such as perception, attention and decision-making increasingly employ multialternative task designs. It is essential in such designs to distinguish the neural correlates of behavioral choices arising from changes in perceptual factors, such as enhanced sensitivity to sensory information, from those arising from changes in decisional factors, such as a stronger bias for a particular response or choice (choice bias). To date such a distinction is not possible with established approaches. Thus, there is a critical need for a theoretical approach that distinguishes the effects of changes in sensitivity from those of changes in choice bias in multialternative tasks. Here, we introduce a mathematical model that decouples choice bias from perceptual sensitivity effects in multialternative detection tasks: multialternative tasks that incorporate catch trials to measure the ability to detect one among multiple (potential) stimuli or stimulus features. By formulating the perceptual decision in a novel, multidimensional signal detection framework, our model identifies the distinct effects of bias and sensitivity on behavioral choices. With a combination of analytical and numerical approaches, we demonstrate that model parameters (sensitivity, bias) are estimated reliably and uniquely, even in tasks involving arbitrarily large numbers of alternatives. Model simulations revealed that ignoring choice bias or performance during catch trials produced systematically inaccurate estimates of perceptual sensitivity, a finding that has important implications for interpreting behavioral data in multialternative detection and cued attention tasks. The model will find important application in identifying the effects of neural perturbations (stimulation or inactivation) on an animal's perception in multialternative attention and decision-making tasks.

q-bio.NC↗