Biology subjects

Lowet, A. S.

Publications and source records attributed to Lowet, A. S..

3 recordsLinked to original sources

bioRxiv2026

The neural circuit implementation of distributional reinforcement learning

Distributional reinforcement learning (RL) algorithms outperform traditional RL by representing the full probability distribution of future outcomes rather than its mean. Neural signatures consistent with distributional coding have been observed in dopamine neurons and in striatal value representations. However, how distributional RL is implemented within the anatomical circuit organization of the basal ganglia remains unknown. Uncovering how representations of outcome distributions are organized across brain circuits has broad implications for both fundamental and clinical neuroscience, providing new insights into how the brain represents uncertainty and how these representations shape risk-sensitive learning and decision-making. Here, we investigated how representations of outcome distributions are organized across dopamine projections to the striatum by combining striatum-wide fiber photometry of dopamine axonal activity with Neuropixels recordings from projection-defined dopamine neurons. We found that distributional reward prediction error signals are preferentially encoded in lateral ventral striatum-projecting dopamine neurons, but that reward distribution information is shared between the ventral and the dorsal striatum. Additionally, an anatomical gradient of optimism in the midbrain supports the diversity of value predictions required for distributional RL. Our findings extend the actor-critic framework proposed for the basal ganglia by revealing distributional prediction errors in a subpopulation of mesolimbic dopamine neurons capable of training parallel striatal regions thought to implement different types of behavioral control.

neuroscience↗

bioRxiv2026

A hippocampal digital Twin reveals latent structure in freely flying bats

Concurrent advances in artificial intelligence (AI), large-scale neuronal recording, and behavioral tracking have enabled increasingly powerful predictive models of brain-behavior relationships. Yet predictive accuracy does not automatically yield interpretable principles of neural organization, particularly during natural behavior, where the relevant variables computed are often unknown in advance. Here, we developed a functional digital twin of the hippocampus, trained jointly on large-scale neural activity and self-directed three-dimensional flight in freely flying bats. A multimodal, multitask transformer accurately predicted both neural activity and behavior, outperformed standard approaches, and generalized across recording sessions. Interrogating the model's representations suggested a form of hippocampal generalization across self-selected trajectories that was not apparent in physical coordinates, but was plainly visible in the model's low-dimensional latent space. We confirmed this organization directly in the recorded neural activity: neurons fired at corresponding stages of spatially distinct paths connecting the same origin and destination, preserving their relative firing order even as their firing fields shifted with the path. Analysis of neuron-specific readouts further identified functional relationships among neurons that generalized from flight to waking rest and sleep, brain states never used for training, and differed systematically between the hippocampal subregions CA1 and CA3. Together, these results show how AI-based functional digital twins can complement conventional analysis by generating testable hypotheses about neural organization during natural behavior.

neuroscience↗

bioRxiv2024

An opponent striatal circuit for distributional reinforcement learning

Machine learning research has achieved large performance gains on a wide range of tasks by expanding the learning target from mean rewards to entire probability distributions of rewards -- an approach known as distributional reinforcement learning (RL)1. The mesolimbic dopamine system is thought to underlie RL in the mammalian brain by updating a representation of mean value in the striatum2,3, but little is known about whether, where, and how neurons in this circuit encode information about higher-order moments of reward distributions4. To fill this gap, we used high-density probes (Neuropixels) to acutely record striatal activity from well-trained, water-restricted mice performing a classical conditioning task in which reward mean, reward variance, and stimulus identity were independently manipulated. In contrast to traditional RL accounts, we found robust evidence for abstract encoding of variance in the striatum. Remarkably, chronic ablation of dopamine inputs disorganized these distributional representations in the striatum without interfering with mean value coding. Two-photon calcium imaging and optogenetics revealed that the two major classes of striatal medium spiny neurons -- D1 and D2 MSNs -- contributed to this code by preferentially encoding the right and left tails of the reward distribution, respectively. We synthesize these findings into a new model of the striatum and mesolimbic dopamine that harnesses the opponency between D1 and D2 MSNs5-15 to reap the computational benefits of distributional RL.

neuroscience↗