Perceptual learning of bisection stimuli under roving: Slow and largely specific

Parkosadze, Khatuna; Otto, T.; Malania, Maka; Kezeli, A.; Herzog, Michael H.

doi:10.1167/8.1.5

Cited by 29 publications

(33 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Randomly mixing easy and difficult stimuli ('roving') in a classification task hampers learning with the 2-component rule, but not with the rate-optimized 1-component rule. A similar phenomenon is observed in perceptual discrimination tasks when bisection stimuli of unequal difficulties are randomly intermixed (Otto et al, 2006;Parkosadze et al, 2008;Clarke et al, 2014). In our model performance degrades because the easy stimuli simultaneously trigger suprathreshold weight changes and suprathreshold voltages that are then prematurely consolidated, even when these changes are not accurate enough to correctly classify the difficult stimuli.…”

Section: Discussionsupporting

confidence: 63%

“…If the easy and difficult stimuli are presented randomly interleaved, the 2-component rule performs less well, unlike the 1-component rule (Fig. 7D), analogously to human behaviour (Parkosadze et al, 2008;Clarke et al, 2014). The performance decrease with roving (mixing) occurs because weight changes, in response to an easy stimulus, point further away from the optimal weight due to the larger variance of presynaptic rates around the mean for these easy stimuli.…”

Section: Consolidation By Selecting Informative Learning Eventsmentioning

confidence: 93%

“…If in a given context classification tasks of unequal difficulties are mixed, learning slows down. This is indeed a well known phenomenon in psychophysics, called roving (Tartaglia et al, 2009b;Parkosadze et al, 2008) or interleaved learning (Flesch et al, 2018).…”

Section: Consolidation By Selecting Informative Learning Eventsmentioning

confidence: 97%

See 2 more Smart Citations

Synaptic weight decay with selective consolidation enables fast learning without catastrophic forgetting

Leimer

Herzog

Senn

2019

Preprint

View full text Add to dashboard Cite

Learning can interfere with pre-existing memories that in classical neural networks may lead to catastrophic forgetting. Different from these networks, biological synapses show an early decay of long-term potentiation, combined with a tag & capture mechanism for selective consolidation. We present a 2component synaptic plasticity model that, by an early decay and a tag & capture mechanism, enables context-dependent fast learning without catastrophic forgetting. For reinforcement learning in multiple contexts, the fast learning combined with an early weight decay maximizes the expected reward while minimizing interferences between subsequent contexts. Fast learning, enabled by a highly plastic weight component, improves performance for a given context. Between contexts this plastic component decays to prevent interference, but selective consolidation into a stable component protects old memories. As a downside of these mechanisms, learning is hampered when consolidation is triggered prematurely by interleaving easy and difficult tasks, consistent with human psychophysical experiments.

show abstract

Section: Discussionsupporting

confidence: 63%

Section: Consolidation By Selecting Informative Learning Eventsmentioning

confidence: 93%

See 1 more Smart Citation

Synaptic weight decay with selective consolidation enables fast learning without catastrophic forgetting

Leimer

Herzog

Senn

2019

Preprint

View full text Add to dashboard Cite

show abstract

“…It is important to note that while these handcrafted weights are not robust to change in the spacings of the outer bars, it is possible to learn the network for different spacings using BPTT algorithm (data not shown). Recently Herzog and his colleagues (Parkosadze et al 2008) have demonstrated that it is possible to learn different spacings.…”

Section: Robustness To Variations In Weight Matrix and Number Of Linementioning

confidence: 99%

Change-based inference for invariant discrimination

Moazzezi

Dayan

2008

Network: Computation in Neural Systems

View full text Add to dashboard Cite

Under a conventional view of information processing in recurrently connected populations of neurons, computations consist in mapping inputs onto terminal attractor states of the dynamical interactions. However, there is evidence that substantial information representation and processing can occur over the course of the initial evolution of the dynamical states of such populations, a possibility that has attractive computational properties. Here, we suggest a model that explores one such property, namely, the invariance to an irrelevant feature dimension that arises from monitoring not the state of the population, but rather (a statistic of) the change in this state over time. We illustrate our proposal in the context of the bisection task, a paradigmatic example of perceptual learning for which an attractor-state recurrent model has previously been suggested. We show a change-based inference scheme that achieves near optimal performance in the task (with invariance to translation), is robust to high levels of dynamical noise and variations of the synaptic weight matrix, and indeed admits a computationally straightforward learning rule.

show abstract

“…In contrast to the alternative model that perceptual learning is based on modifying intrinsic V1 connections [13], it confirms improvements in bisection learning under stimulus roving [12], and a weak learning transfer from a trained to a non-trained stimulus width [11]. It moreover makes several testable predictions both on the behavioral and the neuronal level:…”

Section: Discussionmentioning

confidence: 53%

Adaptive Gain Modulation in V1 Explains Contextual Modifications during Bisection Learning

2009

View full text Add to dashboard Cite

The neuronal processing of visual stimuli in primary visual cortex (V1) can be modified by perceptual training. Training in bisection discrimination, for instance, changes the contextual interactions in V1 elicited by parallel lines. Before training, two parallel lines inhibit their individual V1-responses. After bisection training, inhibition turns into non-symmetric excitation while performing the bisection task. Yet, the receptive field of the V1 neurons evaluated by a single line does not change during task performance. We present a model of recurrent processing in V1 where the neuronal gain can be modulated by a global attentional signal. Perceptual learning mainly consists in strengthening this attentional signal, leading to a more effective gain modulation. The model reproduces both the psychophysical results on bisection learning and the modified contextual interactions observed in V1 during task performance. It makes several predictions, for instance that imagery training should improve the performance, or that a slight stimulus wiggling can strongly affect the representation in V1 while performing the task. We conclude that strengthening a top-down induced gain increase can explain perceptual learning, and that this top-down signal can modify lateral interactions within V1, without significantly changing the classical receptive field of V1 neurons.

show abstract

Perceptual learning of bisection stimuli under roving: Slow and largely specific

Cited by 29 publications

References 30 publications

Synaptic weight decay with selective consolidation enables fast learning without catastrophic forgetting

Synaptic weight decay with selective consolidation enables fast learning without catastrophic forgetting

Change-based inference for invariant discrimination

Adaptive Gain Modulation in V1 Explains Contextual Modifications during Bisection Learning

Contact Info

Product

Resources

About