STATE-DEPENDENT CIRCUIT COMMUNICATION DRIVING ADAPTIVE STRATEGIES

Decision-making is shaped not only by external cues but also by internal states, such as motivation and anxiety. Even in identical situations, animals adopt diverse decision-making strategies, ranging from rigid, stimulus-bound rules to flexible, inference-based behavior. While anxiety-like internal states influence decision-making, the neural mechanisms through which they bias strategy selection remain unclear. Here, we investigate how anxiety-like traits shape strategy use and cognitive flexibility in mice performing a head-fixed virtual foraging task. We hypothesize that individual internal states differentially bias strategy selection, such that higher anxiety-like levels are associated with increased reliance on stimulus-bound strategies. Mice are trained to collect probabilistic water rewards at different locations in a virtual environment and exhibit stable inter-individual differences in strategy use across sessions. Throughout training, anxiety-like traits are characterized by combining a battery of behavioral assays (including the open field, light–dark box, social interaction, and sucrose preference tests) with non-invasive physiological measures of fecal corticosterone level fluctuations. By quantifying behavioral signatures of stimulus-bound and inference-based strategies, ongoing analyses use dimensionality reduction to identify multivariate anxiety-like profiles associated with distinct decision strategies. In parallel, to link behavioral variability to neural activity, we are acquiring and analyzing large-scale electrophysiological recordings using Neuropixels probes in brain regions implicated in strategy selection, including the orbitofrontal cortex, secondary motor cortex, and dorsolateral striatum, during task performance. This work establishes an experimental framework to study how anxiety-like behavioral traits shape strategic diversity, providing a foundation for future circuit-level and causal investigations.