Some context-dependent responses may be maladaptive; for example, anxiety may repress the recollection of information under stressful conditions. Others may be adaptive, enabling discretion and choice in the face of risk or opportunity. More broadly, contextual factors, related to brain state and cost-benefit calculations, weigh on behavioral decisions and impact the interpretation of self-reported knowledge. Thus, it is crucial to disambiguate the effects of context from knowledge when interpreting performance, particularly during learning.

My lab studies the neural circuits and dynamics that enable learning, with an emphasis on the role of context and brain state. We first aim to gain behavioral control of learning-related phenomena using quantitative assessment of mouse behavior. We then apply the modern tools of neuroscience to monitor, manipulate and model neural networks to determine how the brain implements learning-related computations. We have a particular interest in the role of neuromodulation, including cholinergic and noradrenergic, in cortical and subcortical circuits.

We also apply insights from neural circuit research to disease states. In Alzheimer’s disease (AD), for example, familiar contexts can trigger episodes of lucidity even when patients are deep in cognitive decline. AD may impinge on the learning processes we study, but in reverse. Knowledge may exist in the AD brain but become inaccessible. Can this ‘hidden’ knowledge be unlocked? These momentary retrievals may reflect activation of context-dependent mechanisms and could be ripe for therapeutic intervention to improve cognition. More generally, neuroscience research that introduces a robust dialectic between the study of healthy and diseased cognitive states will undoubtedly reveal deeper insights into both.