Stochastic interactive processes and the effect of context on perception.

By James L. McClelland. 1991

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Generalization Through the Recurrent Interaction of Episodic Memories (2012, methodology): “Intrinsic variability also allows recurrent networks to conform to principles of Bayesian computation that may not be perfectly adhered to by the continuous approximation (McClelland, 1991; Mirman, Khaitan, Bolger, & McClelland, in press).”
Connectionist and diffusion models of reaction time. (1999, methodology): “The Interactive Activation Model, McClelland & Rumelhart (1981). The interactive activation model was designed to explain performance in simple letter and word identification tasks.”
The time course of perceptual choice: the leaky, competing accumulator model. (2001, background): “However, McClelland (1991) and Movellan & McClelland (1995) have shown that with intrinsic variability incorporated, the interactive-activation model accounts for the pattern of data in question, demonstrating that Massaro’s claim was incorrect.”
Trace synthesis in cued recall (1992, methodology): “McClelland (1991) indicates that variability may be introduced in a variety of ways with similar results; in the present model, we simply injected a small amount of normally distributed random noise into the input to each unit at each update.”
Is vision continuous with cognition? The case for cognitive impenetrability of visual perception. (1999, methodology): “Similarly, McClelland (1991) used a d' measure to argue in favor of a context effect on phoneme perception and Goldstone (1994) used it to show that discriminability of various dimensions could be changed after learning to categorize stimuli in which those dimensions were relevant.”
Understanding normal and impaired word reading: computational principles in quasi-regular domains. (1996, methodology): “Note that this determinacy is not an intrinsic limitation of connectionist networks (see, e.g., Movellan & McClelland, 1991).”
Interactive Activation and Mutual Constraint Satisfaction in Perception and Cognition (2014, methodology): “While the original model’s assumptions did distort this relationship, the problem was not in fact due to interactivity: As mentioned above, the influence of multiple sources of input failed to exhibit logistic additivity under the activation functions used in the original models even when propagatio”
Précis of Semantic Cognition: A Parallel Distributed Processing Approach (2008, methodology): “Such models have been brought to bear on a broad range of phenomena relating to knowledge about sequential structure (e.g., Cleeremans 1993; Cleeremans & McClelland 1991; Elman 1990; 1991; Rohde & Plaut 1999), including models of language comprehension.”
Connectionist models of development (2003, methodology): “An illustration of this property of simple recurrent networks (the type of network used by Elman) is presented in a simulation performed by Servan-Schreiber, Cleeremans and McClelland (1991).”
Connectionist modelling in psychology: A localist manifesto (2000, methodology): “The illustration is easiest to perform for the case of two-alternative choice, and is the same as that found in McClelland (1991).”
Models of accuracy in repeated-measures designs (2008, background): “One such argument is based on Luce (1963) choice theory (cf. McClelland, 1991).”
Merging information in speech recognition: Feedback is never necessary (2000, background): “…long-term average behaviour of a connectionist network follow the Luce choice rule by adding noise and then selecting the node with the largest activation (Page 2000; McClelland 1991), there is no such obvious way to derive response probabilities directly from network activations on a single trial.”