Spontaneous animal behaviour is built from action modules that are concatenated by the brain into sequences1,2. However, the neural mechanisms that guide the composition of naturalistic, self-motivated behaviour remain unknown. Here we show that dopamine systematically fluctuates in the dorsolateral striatum (DLS) as mice spontaneously express sub-second behavioural modules, despite the absence of task structure, sensory cues or exogenous reward. Photometric recordings and calibrated closed-loop optogenetic manipulations during open field behaviour demonstrate that DLS dopamine fluctuations increase sequence variation over seconds, reinforce the use of associated behavioural modules over minutes, and modulate the vigour with which modules are expressed, without directly influencing movement initiation or moment-to-moment kinematics. Although the reinforcing effects of optogenetic DLS dopamine manipulations vary across behavioural modules and individual mice, these differences are well predicted by observed variation in the relationships between endogenous dopamine and module use. Consistent with the possibility that DLS dopamine fluctuations act as a teaching signal, mice build sequences during exploration as if to maximize dopamine. Together, these findings suggest a model in which the same circuits and computations that govern action choices in structured tasks have a key role in sculpting the content of unconstrained, high-dimensional, spontaneous behaviour.

Keypoint tracking algorithms have revolutionized the analysis of animal behavior, enabling investigators to flexibly quantify behavioral dynamics from conventional video recordings obtained in a wide variety of settings. However, it remains unclear how to parse continuous keypoint data into the modules out of which behavior is organized. This challenge is particularly acute because keypoint data is susceptible to high frequency jitter that clustering algorithms can mistake for transitions between behavioral modules. Here we present keypoint-MoSeq, a machine learning-based platform for identifying behavioral modules ("syllables") from keypoint data without human supervision. Keypoint-MoSeq uses a generative model to distinguish keypoint noise from behavior, enabling it to effectively identify syllables whose boundaries correspond to natural sub-second discontinuities inherent to mouse behavior. Keypoint-MoSeq outperforms commonly-used alternative clustering methods at identifying these transitions, at capturing correlations between neural activity and behavior, and at classifying either solitary or social behaviors in accordance with human annotations. Keypoint-MoSeq therefore renders behavioral syllables and grammar accessible to the many researchers who use standard video to capture animal behavior.

Characterizing animal behavior requires methods to distill 3D movements from video data. Though keypoint tracking has emerged as a widely used solution to this problem, it only provides a limited view of pose, reducing the body of an animal to a sparse set of experimenter-defined points. To more completely capture 3D pose, recent studies have fit 3D mesh models to subjects in image and video data. However, despite the importance of mice as a model organism in neuroscience research, these methods have not been applied to the 3D reconstruction of mouse behavior. Here, we present ArMo, an articulated mesh model of the laboratory mouse, and demonstrate its application to multi-camera recordings of head-fixed mice running on a spherical treadmill. Using an end-to-end gradient based optimization procedure, we fit the shape and pose of a dense 3D mouse model to data-derived keypoint and point cloud observations. The resulting reconstructions capture the shape of the animal's surface while compactly summarizing its movements as a time series of 3D skeletal joint angles. ArMo therefore provides a novel alternative to the sparse representations of pose more commonly used in neuroscience research.