We report a model for object identification based on an exper- iment that varies the arrival times of different features of the objects. A single object, a circle with four spokes extending in different directions, is presented and must be classified as either one of four well trained target stimuli, or one of four well trained foil stimuli. The features (spokes) are presented either simultaneously or successively at intervals of 16, 33, or 50 ms., with target diagnostic features arriving first or last. All durations are short enough that the display appears simultane- ous. The data show that individual decisions vary with both timing and diagnosticity. We apply a dynamic model based on one reported in (Cox & Shiffrin, 2017) for episodic recognition memory. Our model assumes features are perceived at vary- ing times following presentation, possibly in error. At each moment the current features are compared to the well learned memory representations of the eight stimuli, producing a like- lihood ratio for target vs foil. A decision is made when the log likelihood first exceeds a target decision boundary or falls be- low a foil decision boundary. The model implements a form of Bayesian optimal decision making given the assumptions con- cerning feature perception. It predicts the key findings quite well.

The present research infers aspects of spatial attention frommovement to targets (and preferably not to foils) of a mouse-controlled cursor on a computer monitor. The long-term goalis a data-rich and rapid assessment technique that can be usedto diagnose individual and clinical deficits of attention. Theaim of this present research is validating the approach usinga college population of subjects. In the experiment, partici-pants attempt to move a cursor toward three spatial positions atwhich targets appear rapidly but at irregular times, and attemptto inhibit movements toward foils appearing at those positions.We assume that cursor movements toward a position indicatesattention has been directed toward that position. A modifiedHidden Markov Model (HMM) uses five sources of evidence,all based on parameters to be estimated, to predict the timevarying movement of attention: four aspects of cursor move-ment and a probability that attention will move from one timeinterval to the next. Five minutes of data are used to estimateparameters for each subject that produce a predicted attentiontrajectory which best matches what the subject is instructed todo. These parameters are used to predict the attention trajec-tory for the remainder of the hour of testing. The predictionsof attention movements are then matched to the appearance oftargets and foils to infer such components of attention as abil-ity to respond to targets vs foils, times to do so, and changesin these components over time. The results illustrate a promis-ing approach to assessment of attention that could likely beemployed for both adults and children in clinical settings re-quiring short testing periods.

In a world that relies increasingly on large amounts of data and on powerful Machine Learning (ML) models, the veracity of decisions made by these systems is essential. Adversarial samples are inputs that have been perturbed to mislead the in- terpretation of the ML and are a dangerous vulnerability. Our research takes a first step into what can be an important innova- tion in cognitive science: we analyzed human’s judgments and decisions when confronted with targeted (inputs constructed to make a ML model purposely misclassify an input as some- thing else) and non-targeted (a noisy perturbed input that tries to trick the ML model) adversarial samples. Our findings sug- gest that although ML models that produce non-targeted adver- sarial samples can be more efficient than targeted samples they result in more incorrect human classifications than those of tar- geted samples. In other words, non-targeted samples interfered more with human perception and categorization decisions than targeted samples.

Social stimuli are a highly salient source of information, and seem to possess unique qualities that set them apart from other well-known categories. One characteristic is their ability to elicit spatial orienting, whereby directional stimuli like eye- gaze and pointing gestures act as exogenous cues that trigger automatic shifts of attention that are difficult to inhibit. This effect has been extended to non-social stimuli, like arrows, leading to some uncertainty regarding whether spatial orienting is specialized for social cues. Using a standard spatial cueing paradigm, we found evidence that both a pointing hand and arrow are effective cues, but that the hand is encoded more quickly, leading to overall faster responses. We then extended the paradigm to include multiple cues in order to evaluate congruent vs. incongruent cues. Our results indicate that faster encoding of the social cue leads to downstream effects on the allocation of attention resulting in faster orienting.