Purchase this article with an account.
Jiajuan Liu, Zhong-Lin Lu, Barbara Dosher; Augmented Hebbian learning accounts for the Eureka effect in perceptual learning. Journal of Vision 2009;9(8):851. doi: 10.1167/9.8.851.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Previous analyses of the role of feedback have suggested that perceptual learning may be accomplished through augmented Hebbian learning (Petrov, Dosher, & Lu, 2005; 2006) When there is feedback, the product of the feedback and the input is used to update the weights in the neural network; in the absence of feedback, the product of the output and the input is used to update the weights. One prediction of this learning rule is that the ability to exhibit perceptual learning without feedback may depend on the training accuracy level, which has been confirmed in our lab. Another prediction is that the existence of high accuracy trials may facilitate the learning in low accuracy trials. We tested this “Eureka” effect. The accelerated stochastic approximation method was used to track threshold contrasts at particular performance accuracy levels in a Gabor orientation identification task over 6 training days. Subjects were divided into 6 groups: 2 experimental groups in which high training accuracy (85% correct) was mixed with low training accuracy (65% correct) with and without feedback; 4 control groups in which high/low training accuracy was mixed with the same high/low training accuracy with and without feedback. Contrast thresholds improved in the high-high and high-low mixture training accuracy groups independent of the feedback condition. However, threshold improved in the low-low mixture training accuracy condition only in the presence of feedback. Furthermore, the learning rates for both high and low accuracy staircases in a mixture group did not significantly distinguish from each other; nor did they differ from those of control groups. The results are both qualitatively and quantitatively consistent with the predictions of an augmented Hebbian learning model, but not with pure supervised error correction or pure Hebbian learning models. The results lend further support for the augmented Hebbian learning hypothesis in perceptual learning.
This PDF is available to Subscribers Only