Complementary Learning Systems

O’Reilly, Randall C.; Bhattacharyya, Rajan; Howard, Michael D.; Ketz, Nicholas

doi:10.1111/j.1551-6709.2011.01214.x

Cited by 223 publications

(176 citation statements)

References 89 publications

(118 reference statements)

Supporting

Mentioning

164

Contrasting

Order By: Relevance

“…Furthermore, memory retrieval during complex span tasks is associated with increased medial temporal lobe activity (Chein, Moore, & Conway, 2011;Faraco et al, 2011). As the medial temporal lobe is thought to be involved with the binding of material in long-term memory (McClelland, McNaughton, & O'Reilly, 1995;O'Reilly, Bhattacharyya, Howard, & Ketz, 2011), these tasks may rely on temporary storage and selective retrieval from episodic memory rather than (or in addition to) active maintenance.…”

Section: Control Of Attentionmentioning

confidence: 99%

The scope and control of attention: Sources of variance in working memory capacity

Conway

2015

Mem Cogn

View full text Add to dashboard Cite

Working memory capacity is a strong positive predictor of many cognitive abilities, across various domains. The pattern of positive correlations across domains has been interpreted as evidence for a unitary source of inter-individual differences in behavior. However, recent work suggests that there are multiple sources of variance contributing to working memory capacity. The current study (N = 71) investigates individual differences in the scope and control of attention, in addition to the number and resolution of items maintained in working memory. Latent variable analyses indicate that the scope and control of attention reflect independent sources of variance and each account for unique variance in general intelligence. Also, estimates of the number of items maintained in working memory are consistent across tasks and related to general intelligence whereas estimates of resolution are task-dependent and not predictive of intelligence. These results provide insight into the structure of working memory, as well as intelligence, and raise new questions about the distinction between number and resolution in visual short-term memory.Keywords Working memory · Working memory capacity · Individual differences · Visual arrays · Complex span · Intelligence Working memory is a limited capacity system responsible for the active maintenance of information, as well as its M. Chow · A. R. A. Conway ( )

show abstract

Section: Control Of Attentionmentioning

confidence: 99%

The scope and control of attention: Sources of variance in working memory capacity

Conway

2015

Mem Cogn

View full text Add to dashboard Cite

show abstract

“…"), while remembering specific facts and events (e.g., "Where did I put my car keys today?") is processed by the medial temporal lobe, including the hippocampus and surrounding cortical structures (Squire, 2004;Atallah et al, 2004;O'Reilly et al, 2011). The theory and model presented in this letter are aimed at replicating dissociable forms of repetition priming observed in the fusiform and occipital-temporal cortices, but we speculate that the inverted U-shaped dependence of neural activity with respect to the novelty of observed data could be a more general property of the cortex, including deep structures, such as the rhinal, inferior temporal, parahippocampal cortices, and, arguably, the hippocampus.…”

Section: Discussionmentioning

confidence: 99%

“…While our model has some algorithmic similarity to the implementation of the Leabra system (O'Reilly & Munakata, 2000) and its further extension, the complementary learning system (CLS) (O'Reilly et al, 2011), the current research is focused on different goals. In contrast to the CLS, which offers a model of the operation of the MTL memory system, our model explores fundamental neural mechanisms that may be present in various memory areas, giving rise to an inverted-U shape function of neural activation with respect to familiarity.…”

Section: Further Implicationsmentioning

confidence: 99%

“…For example, the model of complementary learning systems (Atallah, Frank, & O'Reilly, 2004;O'Reilly, Bhattacharyya, Howard, & Ketz, 2011), employs a competitive Hebbian learning algorithm that demonstrates cortical activation sharpening after repeated presentation of a stimulus. This algorithm relies on a simulated lateral inhibition mechanism in which strong neurons actively inhibit weaker ones.…”

Section: Models Of Repetition Suppressionmentioning

confidence: 99%

See 1 more Smart Citation

Dissociable Forms of Repetition Priming: A Computational Model

Makukhin

Bolland

2014

Neural Computation

View full text Add to dashboard Cite

Nondeclarative memory and novelty processing in the brain is an actively studied field of neuroscience, and reducing neural activity with repetition of a stimulus (repetition suppression) is a commonly observed phenomenon. Recent findings of an opposite trend-specifically, rising activity for unfamiliar stimuli-question the generality of repetition suppression and stir debate over the underlying neural mechanisms. This letter introduces a theory and computational model that extend existing theories and suggests that both trends are, in principle, the rising and falling parts of an inverted U-shaped dependence of activity with respect to stimulus novelty that may naturally emerge in a neural network with Hebbian learning and lateral inhibition. We further demonstrate that the proposed model is sufficient for the simulation of dissociable forms of repetition priming using real-world stimuli. The results of our simulation also suggest that the novelty of stimuli used in neuroscientific research must be assessed in a particularly cautious way. The potential importance of the inverted-U in stimulus processing and its relationship to the acquisition of knowledge and competencies in humans is also discussed.

show abstract

“…Similarly, the step-like improvements in performance in our training tasks between training days but not within training days suggests that slower offline learning processes (and possibly sleep-specific consolidation processes) are shaping our effects in important ways. Future investigations of these issues could provide insights for theories of complementary memory systems regarding how knowledge of a quasiregular domain is initially learned and gradually integrated with existing knowledge (McClelland, McNaughton, & O'Reilly, 1995;O'Reilly, Bhattacharyya, Howard, & Ketz, 2014).…”

Section: Warping 39mentioning

confidence: 99%

Generalization from newly learned words reveals structural properties of the human reading system.

Armstrong¹,

Dumay²,

Kim³

et al. 2017

Journal of Experimental Psychology: General

View full text Add to dashboard Cite

Connectionist accounts of quasiregular domains, such as spelling-sound correspondences in English, represent exception words (e.g., pint) amidst regular words (e.g., mint) via a graded "warping" mechanism. Warping allows the model to extend the dominant pronunciation to nonwords (regularization) with minimal interference (spillover) from the exceptions. We tested for a behavioral marker of warping by investigating the degree to which participants generalized from newly learned made-up words, which ranged from sharing the dominant pronunciation (regulars), a subordinate pronunciation (ambiguous), or a previously non-existent (exception) pronunciation. The new words were learned over two days, and generalization was assessed 48 hours later using nonword neighbors of the new words in a tempo naming task. The frequency of regularization (a measure of generalization) was directly related to degree of warping required to learn the pronunciation of the new word. Simulations using the Plaut et al. (1996) model further support a warping interpretation. Our findings highlight the need to develop theories of representation that are integrally tied to how those representations are learned and generalized.Keywords: quasiregularity; connectionist models; word learning; tempo naming. Warping 3Generalization from newly learned words reveals structural properties of the human reading system Mastery of reading in alphabetic languages, such as English, requires an individual to learn correspondences between strings of letters and their pronunciations. Typically, this mapping is consistent, whereby a string of letters is pronounced in one way across many words (int in mint, hint, print). Although such regularities can simplify reading acquisition, there are also exceptions that are inconsistent with these regularities (e.g., pint) that must also be learned, and these clearly pose a challenge. Nonetheless, most people become proficient readers. Why does learning an exception word like pint not disrupt reading orthographically similar words like mint? This paper investigates how skilled readers represent both regular words and exceptions to the rule, and yet exhibit only minimal interference between these types of items.Considerable computational works has been devoted to understanding how quasiregularity is represented in memory in a way that enables accurate reading of both regulars and exceptions. For example, the DRC model (Coltheart, Rastle, Perry, Langdon, & Ziegler, 2001) has maintained that two qualitatively different mechanisms are required to do so: a rulebased system to deal with regulars and new words, and a memory-based system to accommodate exceptions. Others have argued that it is possible for a single mechanism to represent both regularities and the various degrees of inconsistency characteristic of natural language (e.g., Plaut, Seidenberg, Patterson, & McClelland, 1996). A connectionist, parallel distributed processing (henceforth PDP) network that maps between spelling and sound via an intermediate pool ...

show abstract

Complementary Learning Systems

Cited by 223 publications

References 89 publications

The scope and control of attention: Sources of variance in working memory capacity

The scope and control of attention: Sources of variance in working memory capacity

Dissociable Forms of Repetition Priming: A Computational Model

Generalization from newly learned words reveals structural properties of the human reading system.

Contact Info

Product

Resources

About