From recency to central tendency biases in working memory: a unifying network model

The central tendency bias, or contraction bias, is a phenomenon where the judgment of the magnitude of items held in working memory (WM) is biased towards the average of past observations. It is assumed to be an optimal strategy by the brain, and commonly thought of an expression of the brain’s ability to learn the statistical structure of sensory inputs. On the other hand, recency biases such as serial dependence are also commonly observed, and are thought to reflect the content of WM. Recent results from an auditory delayed comparison task in rats, suggest that both biases may be more related
than previously thought: when the posterior parietal cortex (PPC) was silenced, both short-term and contraction biases were reduced. Here, we show how a volatile WM content susceptible to shifting to the past sensory experience – producing short-term sensory history biases – naturally leads to contraction bias. Errors at individual trials are sampled from the full distribution of the stimuli, and are not due to a gradual shift of the memory towards the sensory distribution’s mean. Our results are consistent with a broad set of behavioural findings in humans and rodents. We make specific predictions on neuronal dynamics in the PPC and downstream working memory areas, as well as how contraction bias may be altered, upon manipulations of the sensory stimulus distribution, inter-trial and inter-stimulus delay intervals. We also contrast our model predictions for a variety of sensory stimulus distributions against the Bayesian optimal model.