With repeated exposure, people become better at identifying presented items and better at rejecting items that have not been presented. This differentiation effect is captured in a model consisting of item detectors that learn estimates of conditional probabilities of item features. The model is used to account for a number of findings in the recognition memory literature, including (a) the basic differentiation effect (strength-mirror effect), (b) the fact that adding items to a list reduces recognition accuracy (list-length effect) but extra study of some items does not reduce recognition accuracy for other items (null list-strength effect), (c) nonlinear effects of strengthening items on false recognition of similar distractors, (d) a number of different kinds of mirror effects, (e) appropriate z-ROC curves, and (f) one type of deviation from optimality exhibited in recognition experiments.
The authors present the results of their analysis of an auto-associator for use with sparse representations. Their recognition model using it exhibits a list-length effect but no list-strength effect, a dissociation that current models have difficulty producing. Data on the effects of similarity and strengthening that indicate a dissociation between recognition and frequency judgments are also addressed. Receiver operating characteristic curves for the model have slopes between 0.5 and 1.0 and achieve this ratio in a novel way. The model can also predict latencies naturally. The authors' cued-recall model uses an architecture similar to that of the recognition model and where applicable the same parameters. It predicts appropriate amounts of retroactive interference, and analysis reveals an output competition process that relies on distributed representations and has not been proposed before. In this article a kind of neural network called an autoassociator is presented, analyzed, and used as the central part of new models of single-item recognition and cued recall. We were led to investigate this network by the perceived need for a clean-up mechanism in recall models using distributed representations (Goebel & Lewandowsky, 1991; Lewandowsky & Murdock, 1989). We explain this next, before introducing the paradigms that we model. Many current models of human memory use vectors to represent items. One way to conceptualize this is to assume that each component of the vector indicates the degree to which a particular feature forms a part of the item. An additional assumption that may be made is that a given component may simultaneously take significant values in the representations of a number of items, and this can be conceptualized as being due to the fact that items have features in common. When this is the case, the representations are said to be distributed (van Gelder, 1991).
We report data from eight participants who made alignment judgements between a moving object and a stationary, continuously visible 'landmark'. A reversing object had to overshoot the landmark by a significant amount in order to appear to reverse aligned with it. In addition, an adjacent flash irrelevant to the judgment task reliably increased this illusory 'foreshortening'. This and other results are most simply explained by a model in which the flash causes attentional capture, complemented by processes of temporal integration, or backward inhibition, and object representation. A flash used to probe the perception of a moving object's position disrupts that very perception.
A dual-task paradigm was used to examine the effect of withdrawing attentional and/or cognitive resources from the flash-lag judgment. The flash-lag illusion was larger, and performance in a detection task was generally poorer, under dual-task conditions than in single-task control conditions. These effects were particularly pronounced when decisions in the two tasks were required simultaneously, as compared to when they could be made sequentially. The results suggest that a time-consuming process is involved in the flash-lag decision, of such a nature that prolonging the process increases the magnitude of the illusion.
Planning on the 4-disk version of the Tower of London (TOL4) was examined in stroke patients and unimpaired controls. Overall TOL4 solution scores indicated impaired planning in the frontal stroke but not non-frontal stroke patients. Consistent with the claim that processing the relations between current states, intermediate states, and goal states is a key process in planning, the domain-general relational complexity metric was a good indicator of the experienced difficulty of TOL4 problems. The relational complexity metric shared variance with task-specific metrics of moves to solution and search depth. Frontal stroke patients showed impaired planning compared to controls on problems at all three complexity levels, but at only two of the three levels of moves to solution, search depth and goal ambiguity. Non-frontal stroke patients showed impaired planning only on the most difficult quaternary-relational and high search depth problems. An independent measure of relational processing (viz., Latin square task) predicted TOL4 solution scores after controlling for stroke status and location, and executive processing (Trail Making Test). The findings suggest that planning involves a domain-general capacity for relational processing that depends on the frontal brain regions.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.