Super-optimality and relative distance coding in location memory.

The prevailing model of landmark integration in location memory is Maximum Likelihood Estimation, which assumes that each landmark implies a target location distribution that is narrower for more reliable landmarks. This model assumes weighted linear combination of landmarks and predicts that, given optimal integration, the reliability with multiple landmarks is the sum of the reliabilities with the individual landmarks. Super-optimality is reliability with multiple landmarks exceeding optimal reliability given the reliability with each landmark alone; this is shown when performance exceeds predicted optimal performance, found by aggregating reliability values with single landmarks. Past studies claiming super-optimality have provided arguably impure measures of performance with single landmarks given that multiple landmarks were presented at study in conditions with a single landmark at test, disrupting encoding specificity and thereby leading to underestimation in predicted optimal performance. This study, unlike those prior studies, only presented a single landmark at study and the same landmark at test in single landmark trials, showing super-optimality conclusively. Given that super-optimal information integration occurs, emergent information, that is, information only available with multiple landmarks, must be used. With the target and landmarks all in a line, as throughout this study, relative distance is the only emergent information available. Use of relative distance was confirmed here by finding that, when both landmarks are left of the target at study, the target is remembered further right of its true location the further left the left landmark is moved from study to test.

Benefits of multinomial processing tree models with discrete and continuous variables in memory research: an alternative modeling proposal to Juola et al. (2019).

Signal detection theory (SDT) and two-high threshold models (2HT) are often used to analyze accuracy data in recognition memory paradigms. However, when reaction times (RTs) and/or confidence levels (CLs) are also measured, they usually are analyzed separately or not at all as dependent variables (DVs). We propose a new approach to include these variables based on multinomial processing tree models for discrete and continuous variables (MPT-DC) with the aim to compare fits of SDT and 2HT models. Using Juola et al.'s (2019, Memory & Cognition, 47[4], 855-876) data we have found that including CLs and RTs reduces the standard errors of parameter estimates and accounts for interactions among accuracy, CLs, and RTs that classical versions of SDT and 2HT models do not. In addition, according to the simulations, there is an increase in the proportion of correct model selections when relevant DV are included. We highlight the methodological and substantive advantages of MPT-DC in the disentanglement of contributing processes in recognition memory.

Recognition memory decisions made with short- and long-term retrieval.

In the present research, we produce a coherent account of the storage and retrieval processes in short- and long-term event memory, and long-term knowledge, that produce response accuracy and response time in a wide variety of conditions in our studies of recognition memory. Two to nine pictures are studied sequentially followed by a target or foil test picture in four conditions used in Nosofsky et al. Journal of Experimental Psychology: Learning, Memory, and Cognition, 47, 316-342, (2021) and in our new paradigm: VM: target and foil responses to a given stimulus change from trial to trial; CM: the responses do not change from trial to trial; AN: every trial uses new stimuli; MIXED: combinations of VM, CN, and AN occur on each trial. In the new paradigm a given picture is equally often tested as old or new, but only in CM is the response key the same and learnable. Our model has components that have appeared in a variety of prior accounts, including learning and familiarity, but are given support by our demonstration that accuracy and response time data from a large variety of conditions can be predicted by these processes acting together, with parameter values that largely are unchanged. A longer version of this article, containing information not found here due to space, is available online  https://doi.org/10.31234/osf.io/h8msp .The avalibility of the data (supplement materials), info and link is attached at the end section ( https://psyarxiv.com/h8msp .).

The Impact of Sparse Coding on Memory Lifetimes in Simple and Complex Models of Synaptic Plasticity.

Models of associative memory with discrete state synapses learn new memories by forgetting old ones. In the simplest models, memories are forgotten exponentially quickly. Sparse population coding ameliorates this problem, as do complex models of synaptic plasticity that posit internal synaptic states, giving rise to synaptic metaplasticity. We examine memory lifetimes in both simple and complex models of synaptic plasticity with sparse coding. We consider our own integrative, filter-based model of synaptic plasticity, and examine the cascade and serial synapse models for comparison. We explore memory lifetimes at both the single-neuron and the population level, allowing for spontaneous activity. Memory lifetimes are defined using either a signal-to-noise ratio (SNR) approach or a first passage time (FPT) method, although we use the latter only for simple models at the single-neuron level. All studied models exhibit a decrease in the optimal single-neuron SNR memory lifetime, optimised with respect to sparseness, as the probability of synaptic updates decreases or, equivalently, as synaptic complexity increases. This holds regardless of spontaneous activity levels. In contrast, at the population level, even a low but nonzero level of spontaneous activity is critical in facilitating an increase in optimal SNR memory lifetimes with increasing synaptic complexity, but only in filter and serial models. However, SNR memory lifetimes are valid only in an asymptotic regime in which a mean field approximation is valid. By considering FPT memory lifetimes, we find that this asymptotic regime is not satisfied for very sparse coding, violating the conditions for the optimisation of single-perceptron SNR memory lifetimes with respect to sparseness. Similar violations are also expected for complex models of synaptic plasticity.

Specifying a relationship between semantic and episodic memory in the computation of a feature-based familiarity signal using MINERVA 2.

Approaches to modeling episodic recognition memory often imply a separability from semantic memory insofar as an implicit tabula rasa (i.e., blank slate) assumption is apparent in many simulations. This is evident in the common practice of having new test probes correspond to zero memory traces in the store while old test probes correspond to traces representing instances of items' occurrence on a study list. However, in list-learning studies involving word lists, none of the test items would actually correspond to zero items in the person's memory, as all of the test words are generally known to participants, whether old or new. By focusing on a list-learning recognition phenomenon that likely results from feature-based familiarity detection and necessarily involves a role of preexisting knowledge in its mechanisms-the semantic-feature-based recognition without cued recall phenomenon-we show how incorporating preexisting knowledge into the MINERVA 2 model enables it to simulate previously shown empirical patterns with this phenomenon. The simulation patterns reported here raise new theoretical implications worth further exploration, such as the extent to which the variances change in the signal versus the noise distribution when preexisting knowledge is present versus absent in the simulations.

Brain-inspired automated visual object discovery and detection.

Despite significant recent progress, machine vision systems lag considerably behind their biological counterparts in performance, scalability, and robustness. A distinctive hallmark of the brain is its ability to automatically discover and model objects, at multiscale resolutions, from repeated exposures to unlabeled contextual data and then to be able to robustly detect the learned objects under various nonideal circumstances, such as partial occlusion and different view angles. Replication of such capabilities in a machine would require three key ingredients: (i) access to large-scale perceptual data of the kind that humans experience, (ii) flexible representations of objects, and (iii) an efficient unsupervised learning algorithm. The Internet fortunately provides unprecedented access to vast amounts of visual data. This paper leverages the availability of such data to develop a scalable framework for unsupervised learning of object prototypes-brain-inspired flexible, scale, and shift invariant representations of deformable objects (e.g., humans, motorcycles, cars, airplanes) comprised of parts, their different configurations and views, and their spatial relationships. Computationally, the object prototypes are represented as geometric associative networks using probabilistic constructs such as Markov random fields. We apply our framework to various datasets and show that our approach is computationally scalable and can construct accurate and operational part-aware object models much more efficiently than in much of the recent computer vision literature. We also present efficient algorithms for detection and localization in new scenes of objects and their partial views.

Computational Models of Memory Search.

The capacity to search memory for events learned in a particular context stands as one of the most remarkable feats of the human brain. How is memory search accomplished? First, I review the central ideas investigated by theorists developing models of memory. Then, I review select benchmark findings concerning memory search and analyze two influential computational approaches to modeling memory search: dual-store theory and retrieved context theory. Finally, I discuss the key theoretical ideas that have emerged from these modeling studies and the open questions that need to be answered by future research.

Modeling list-strength and spacing effects using version 3 of the retrieving effectively from memory (REM.3) model and its superimposition-of-similar-images assumption.

Shiffrin and Steyvers (1997) introduced a model of recognition memory called retrieving effectively from memory (REM) and successfully applied it to a number of basic memory phenomena. REM incorporates differentiation, wherein item repetitions are accumulated in a single mnemonic trace rather than separate traces. This allows REM to account for several benchmark findings, including the null list-strength effect in recognition (Ratcliff, Clark, & Shiffrin, 1990). The original REM treated massed and spaced repetitions identically, which prevents it from predicting a mnemonic advantage for spaced over massed repetitions (i.e., the spacing effect). However, Shiffrin and Steyvers discussed the possibility that repetitions might be represented in a single trace only if the subject identifies that the repeated item was previously studied. It is quite plausible that subjects would notice repetitions more for massed than for spaced items. Here we show that incorporating this idea allows REM to predict three important findings in the recognition memory literature: (1) the spacing effect, (2) the finding of slightly positive list-strength effects with spaced repetitions, as opposed to massed repetitions or increased study time, and (3) list-strength effects that have been observed using very large strong-to-weak ratios (see Norman, 2002).

A single, simple, statistical mechanism explains resource distribution and temporal updating in visual short-term memory.

Investigations into the way that information is held and integrated within the visual system provides some basis for understanding how visual information is represented and processed. Just over sixty years ago, Swets, Shipley, McKey, and Green (1959) demonstrated that performance within an auditory detection task increases as a function of the square root of the number of stimulus observation intervals, following the predictions of basic sampling theory, indicating the efficient perceptual integration of stimulus information. This principle of observer performance contingent on a constant rate of stimulus sampling also forms the basis of the sample-size model (Palmer, 1990; Sewell, Lilburn, & Smith, 2014) which seeks to provide an account of how memory resources might be divided among item representations in visual short-term memory (VSTM). In this article, we combine the multiple observations paradigm of Swets and colleagues with the VSTM paradigm of Sewell and colleagues and show that the sample-size relationship accounts for both the increase in performance with the number of presentation intervals and the way that performance changes as a function of the number of items in memory. The model provides an account of both the overall information limit of VSTM and an account of the dynamics of that limit, demonstrating not only that observers can selectively update specific representations in memory but that performance in this task is accounted for by a simple statistical constraint. We discuss the implications for models of VSTM capacity and architecture generally, focusing on the implications for objecthood and the characteristics of encoding to and retrieval from memory.

Mechanisms of output interference in cued recall.

The primary aim of this paper is to elucidate the mechanisms governing output interference in cued recall. Output interference describes the phenomenon where accuracy decrease over the course of an episodic memory test. Output inference in cued recall takes the form of a decrease in correct and intrusion responses and an increase in failures to response across the test. This pattern can only be accounted for by a model with two complementary mechanisms: learning during retrieval and a response filter that prevents repeated recall of the same item. We investigate how a retrieval filter might operate by manipulating the similarity of words. The data are consistent with a retrieval filter that does not operate by a global match of a potential target to previously recalled items. Results are discussed within the search of associative memory theory.

Visual similarity effects in immediate serial recall and (sometimes) in immediate serial recognition.

Words that sound dissimilar are recalled better than otherwise comparable words that sound similar on both immediate serial recall and immediate serial recognition tests, the so-called acoustic similarity effect. Although studies using immediate serial recall have shown an analogous visual similarity effect, in which words that look dissimilar are recalled better than words that look similar, this effect has not been examined in immediate serial recognition. We derived a prediction from the Feature Model that a visual similarity effect will be observed in immediate serial recognition only when the items are acoustically dissimilar; the model predicts no effect when the items are acoustically similar. Experiments 1 and 2 used visually dissimilar and visually similar stimuli that were all acoustically similar and replicated the visual similarity effect in serial recall but revealed no effect in serial recognition. Experiments 3 and 4 used a second set of stimuli that were acoustically dissimilar and found a visual similarity effect in both serial recall and serial recognition. The experiments confirm the Feature Model's predictions and add to earlier findings that the two tests, serial recall and serial recognition, may show quite different results because the two tests are not as similar as previously thought.

Temporal grouping and direction of serial recall.

When lists are presented with temporal pauses between groups of items, participants' response times reiterate those pauses. Accuracy is also increased, especially at particular serial positions. By comparing forward with backward serial recall, we tested whether the influence of temporal grouping is primarily a function of serial position or output position. Results favored the latter, both when recall direction was known to participants prior to (Experiment 2) or only after (Experiment 2) studying each list. Alongside fits of variants of a temporal distinctiveness-based model, our findings suggest that the influence of temporal grouping is not just a consequence of grouping information stored during the study phase. Rather, it critically depends on participants cueing with within-chunk position during recall, combined with response suppression.

Familiarity, recollection, and receiver-operating characteristic (ROC) curves in recognition memory.

The Atkinson-Shiffrin theory describes and explains some of the processes involved in storing and retrieving information in human memory. Here we examine predictions of related models for search and decision processes in recognizing information in long-term memory. In some models, recognition is presumably based on a test item's familiarity judgment, and subsequent decisions follow from the sensitivity and decision parameters of signal detection theory. Other models dispense with the continuous notion of familiarity and base recognition on discrete internal states such as relative certainty that an item has or has not been previously studied, with an intermediate state of uncertainty that produces guesses. Still others are hybrid models with two criteria located along a familiarity continuum defining areas for rapid decisions based on high or low familiarities. For intermediate familiarity values, the decision can be delayed pending the results of search for, and occasional recollection of, relevant episodic information. Here we present the results from a study of human recognition memory for lists of words using both response time and error data to construct receiver-operating characteristic (ROC) curves derived from three standard methods based on the same data set. Models are evaluated against, and parameters estimated from, group as well as individual subjects' behavior. We report substantially different ROC curves when they are based on variations in target-word frequency, confidence judgments, and response latencies. The results indicate that individual versus group data must be used with caution in determining the appropriate theoretical interpretation of recognition memory performance.

Dissociating visuo-spatial and verbal working memory: It's all in the features.

Echoing many of the themes of the seminal work of Atkinson and Shiffrin (The Psychology of Learning and Motivation, 2; 89-195, 1968), this paper uses the feature model (Nairne, Memory & Cognition, 16, 343-352, 1988; Nairne, Memory & Cognition, 18; 251-269, 1990; Neath & Nairne, Psychonomic Bulletin & Review, 2; 429-441, 1995) to account for performance in working-memory tasks. The Brooks verbal and visuo-spatial matrix tasks were performed alone, with articulatory suppression, or with a spatial suppression task; the results produced the expected dissociation. We used approximate Bayesian computation techniques to fit the feature model to the data and showed that the similarity-based interference process implemented in the model accounted for the data patterns well. We then fit the model to data from Guérard and Tremblay (2008, Journal of Experimental Psychology: Learning, Memory, and Cognition, 34, 556-569); the latter study produced a double dissociation while calling upon more typical order reconstruction tasks. Again, the model performed well. The findings show that a double dissociation can be modelled without appealing to separate systems for verbal and visuo-spatial processing. The latter findings are significant as the feature model had not been used to model this type of dissociation before; importantly, this is also the first time the model is quantitatively fit to data. For the demonstration provided here, modularity was unnecessary if two assumptions were made: (1) the main difference between spatial and verbal working-memory tasks is the features that are encoded; (2) secondary tasks selectively interfere with primary tasks to the extent that both tasks involve similar features. It is argued that a feature-based view is more parsimonious (see Morey, 2018, Psychological Bulletin, 144, 849-883) and offers flexibility in accounting for multiple benchmark effects in the field.

Central tendency representation and exemplar matching in visual short-term memory.

I address a recent extension of the generalized context model (GCM), a model which excludes prototypes, to the visual short-term memory (VSTM) literature, which is currently deluged with prototype effects. The paper includes a brief review whose aim is to discuss the background and key findings suggesting that prototypes have an obligatory influence on visual short-term memory responses in the same VSTM task that the GCM's random walk extension, EBRW, was extended to account for: Sternberg scanning. I present a new model that incorporates such "central tendency representations" in memory, as well as several other regularities of the literature, and compare its prediction and postdictions to those of the GCM on some unpublished Sternberg scanning data. The GCM cannot account for the pattern in those data without post hoc modifications but the pattern is predicted nicely by the central tendency representation model. Although the new model is certainly wrong, the review and modeling exercise suggest a reconsideration of prototype models may be warranted, at least in the VSTM literature.

50 years of research sparked by Atkinson and Shiffrin (1968).

In this article we review the framework proposed in 1968 by Atkinson and Shiffrin. We discuss the prior context that led to its production, including the advent of cognitive and mathematical modeling, its principal concepts, the subsequent refinements and elaborations that followed, and the way that the framework influenced other researchers to test the ideas and, in some cases, propose alternatives. The article illustrates the large amount of research and the large number of memory models that were directly influenced by this chapter over the past 50 years.

Visual short-term memory capacity predicts the "bandwidth" of visual long-term memory encoding.

We are capable of storing a virtually infinite amount of visual information in visual long-term memory (VLTM) storage. At the same time, the amount of visual information we can encode and maintain in visual short-term memory (VSTM) at a given time is severely limited. How do these two memory systems interact to accumulate vast amount of VLTM? In this series of experiments, we exploited interindividual and intraindividual differences VSTM capacity to examine the direct involvement of VSTM in determining the encoding rate (or "bandwidth") of VLTM. Here, we found that the amount of visual information encoded into VSTM at a given moment (i.e., VSTM capacity), but neither the maintenance duration nor the test process, predicts the effective encoding "bandwidth" of VLTM.

Semantic memories prime autobiographical memories: General implications and implications for everyday autobiographical remembering.

This study investigated the idea that semantic memory activation causes the activation of associated autobiographical memories (e.g., reading the word summer activates knowledge representations in semantic memory, as well as associated personal memories about summer in autobiographical memory). We tested this semantic-autobiographical memory priming hypothesis in three experiments. In Experiment 1, participants were primed with concepts (e.g., summer) on a familiarity task and were then given a word-cue voluntary autobiographical memory task. In support of the hypothesis, the results showed that primed participants had more autobiographical memories overlapping with the primed concepts than control participants. In Experiment 2, participants were similarly primed, but in this case they were given a measure of involuntary autobiographical memory (i.e., Schlagman and Kvavilashvili's (Memory & Cognition, 36, 920-932, 2008) vigilance task). The results of this experiment also supported the semantic-autobiographical memory-priming hypothesis. Experiment 3 ruled out an alternative possibility (i.e., that autobiographical memory processing had occurred in the word familiarity task) by showing that semantic-autobiographical priming had resulted from a priming task (lexical decision) where autobiographical memory processing was unlikely. The implications of these findings are discussed.

The speed of memory errors shows the influence of misleading information: Testing the diffusion model and discrete-state models.

In this report, we evaluate single-item and forced-choice recognition memory for the same items and use the resulting accuracy and reaction time data to test the predictions of discrete-state and continuous models. For the single-item trials, participants saw a word and indicated whether or not it was studied on a previous list. The forced-choice trials had one studied and one non-studied word that both appeared in the earlier single-item trials and both received the same response. Thus, forced-choice trials always had one word with a previous correct response and one with a previous error. Participants were asked to select the studied word regardless of whether they previously called both words "studied" or "not studied." The diffusion model predicts that forced-choice accuracy should be lower when the word with a previous error had a fast versus a slow single-item RT, because fast errors are associated with more compelling misleading memory retrieval. The two-high-threshold (2HT) model does not share this prediction because all errors are guesses, so error RT is not related to memory strength. A low-threshold version of the discrete state approach predicts an effect similar to the diffusion model, because errors are a mixture of responses based on misleading retrieval and guesses, and the guesses should tend to be slower. Results showed that faster single-trial errors were associated with lower forced-choice accuracy, as predicted by the diffusion and low-threshold models.

Multinomial models reveal deficits of two distinct controlled retrieval processes in aging and very mild Alzheimer disease.

Dual-process models of episodic retrieval reveal consistent deficits of controlled recollection in aging and Alzheimer disease (AD). In contrast, automatic familiarity is relatively spared. We extend standard dual-process models by showing the importance of a third capture process. Capture produces a failure to attempt recollection, which might reflect a distinct error from an inability to recollect when attempted (Jacoby et al. Journal of Experimental Psychology: General, 134(2), 131-148, 2005a). We used multinomial process tree (MPT) modeling to estimate controlled recollection and capture processes, as well as automatic retrieval processes, in a large group of middle-aged to older adults who were cognitively normal (N = 519) or diagnosed with the earliest detectable stage of AD (N = 107). Participants incidentally encoded word pairs (e.g., knee bone). At retrieval, participants completed cued word fragments (e.g., knee b_n_) with primes that were congruent (e.g., bone), incongruent (e.g., bend), or neutral (i.e., &&&) to the target (e.g., bone). MPT models estimated retrieval processes both at the group and the individual levels. A capture parameter was necessary to fit MPT models to the observed data, suggesting that dual-process models of this task can be contaminated by a capture process. In both group- and individual-level analyses, aging and very mild AD were associated with increased susceptibility to capture, decreased recollection, and no differences in automatic influences. These results suggest that it is important to consider two distinct modes of attentional control when modeling retrieval processes. Both forms of control (recollection and avoiding capture) are particularly sensitive to cognitive decline in aging and early-stage AD.