Beibei Dai

Intelligent behaviour emerges not solely from neural computation, but from the tight coupling between brain dynamics, body morphology, and environmental structure. This project investigates how ring attractor networks encode spatial goals, integrate sensory inputs, and resolve competing alternatives during movement in complex, real-world environments. Combining high-resolution behavioural tracking, immersive virtual reality for animals, and mechanistic computational modelling, the work will test predictions from the geometry of decision-making framework, in which multi-choice problems are resolved through spatial bifurcations arising from interactions between neural activity, motion, and environmental geometry under varying levels of noise. Linking empirical behavioural trajectories to low-dimensional attractor models will reveal how biological systems achieve robust, noise-tolerant decisions with minimal computational overhead. These principles will be translated to robotic systems to test whether physically grounded, ring-attractor-based control enables scalable, data-efficient coordination and navigation, providing a unified account of intelligence across brain-body-environment systems.