| |
Abstract:
Abstract: Neurophysiological evidence suggests the
hippocampus self-regulates septal acetylcholine release in response
to novel stimuli (Hasselmo & Schnell, 1994). Gluck & Myers
(1993) modeled the hippocampus as an autoencoder. Myers et al.
(1996) argued that the cholinergic input from medial septum
modulates learning rate in this autoencoder. The current extension
of the model implements the cholinergic self-regulation loop. In
the psychological literature, not only stimulus novelty, but also
stimulus salience is considered responsible for various classical
conditioning effects. In a conditioning context, stimulus salience
is the predictive power of a stimulus, whereas stimulus novelty is
the uniqueness of a stimulus. Two competing accounts have been
provided incorporating either the effects of salience
(Rescorla-Wagner, 1972) or novelty (Pearce-Hall, 1980). Here we
will show that at a neurophysiological level these accounts need
not be conflicting. Rather, both stimulus novelty and salience have
the same effect -increasing acetylcholine levels-, increasing
neural plasticity. Acquisition of a conditioned response is
enhanced for both highly salient and novel stimuli, as can be shown
by latent inhibition results. Blocking, specifically one-trial
blocking, can shown to be critically dependent on both novelty and
salience. Failure to produce blocking under systemic acetylcholine
attribution is shown to be equal to failure to detect a stimulus as
either novel or salient.
|
