Abstract
Monkeys can learn the implied ranking of pairs of images drawn from an ordered set, despite never seeing all of the images simultaneously and without explicit spatial or temporal cues. We recorded the activity of posterior parietal cortex (including lateral intraparietal area LIP) neurons while monkeys learned 7-item transitive inference (TI) lists with 2 items presented on each trial. Behavior and neuronal activity were significantly influenced by the ordinal relationship of the stimulus pairs, specifically symbolic distance (the difference in rank) and joint rank (the sum of the ranks). Symbolic distance strongly predicted decision accuracy and learning rate. An effect of joint rank on performance was found nested within the symbolic distance effect. Across the population of neurons, there was significant modulation of firing correlated with the relative ranks of the two stimuli presented on each trial. Neurons exhibited selectivity for stimulus rank during learning, but not before or after. The observed behavior is poorly explained by associative or reward mechanisms, and appears more consistent with a mental workspace model in which implied serial order is mapped within a spatial framework. The neural data suggest that posterior parietal cortex supports serial learning by representing information about the ordinal relationship of the stimuli presented during a given trial.
Highlights
Transitive inference (TI) refers to judgments that if A > B and B > C, one can infer that A > C by the transitive property of ordinal rank[1]
A hallmark of model-free reinforcement learning (RL) is that choices that are rewarded are more likely to be repeated than those that are not. We investigated this by first looking at the pattern of spatial choices; if the subject was rewarded for choosing the stimulus in a particular location, would they be more or less likely to choose the same location on the trial regardless of whether it was the correct response? There was evidence in favor of a spatial win-stay, lose-shift bias; rewarded responses were repeated 56% of the time, significantly different from the expectation of 50% (t-test p < 0.0001, N = 141, Cohen’s d = 0.68)
Much work in humans and other animals has supported the idea that transitive inference (TI) learning relies on a mental workspace, but few studies have examined its physiological underpinnings in brain regions implicated in spatial cognition (e.g.55.)
Summary
Transitive inference (TI) refers to judgments that if A > B and B > C, one can infer that A > C by the transitive property of ordinal rank[1]. Several lines of evidence suggest that abstract cognitive mechanisms give rise to TI abilities[10,11], distinguishing them from more concrete forms of learning such as motor sequences The strongest of these is the presence of a symbolic distance effect (SDE12). Evidence that the firing rates of LIP neurons correlate with ordinal position of or symbolic distance between stimuli, would provide correlational evidence for the encoding of a mental number line in LIP. Such results would support future attempts to validate a workspace for ordinal reasoning of this kind, using causal manipulations such as chemical inactivation
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have