This study investigates the brain-wide dynamics linking sensation to action during decision-making in mice. The researchers used dense silicon electrode recordings to capture neural activity in 51 brain regions while mice performed a visual change detection task. They found that sensory evidence is integrated across widely distributed multi-regional premotor circuits, driven by sparse neural populations that integrate visual input and prepare movements. These integrated representations are learned and emerge with training, enabling mice to make decisions based on ambiguous sensory information. The study reveals that evidence integration and movement preparation are encoded in the same subspace of population activity, orthogonal to movement-related dynamics. Preparatory activity, which resides in this subspace, is initiated by neurons integrating sensory evidence and collapses at movement onset, allowing the integration process to reset. This work provides a unified framework for understanding how sensory evidence controls actions through global neural mechanisms, linking concepts from decision-making and motor control fields.This study investigates the brain-wide dynamics linking sensation to action during decision-making in mice. The researchers used dense silicon electrode recordings to capture neural activity in 51 brain regions while mice performed a visual change detection task. They found that sensory evidence is integrated across widely distributed multi-regional premotor circuits, driven by sparse neural populations that integrate visual input and prepare movements. These integrated representations are learned and emerge with training, enabling mice to make decisions based on ambiguous sensory information. The study reveals that evidence integration and movement preparation are encoded in the same subspace of population activity, orthogonal to movement-related dynamics. Preparatory activity, which resides in this subspace, is initiated by neurons integrating sensory evidence and collapses at movement onset, allowing the integration process to reset. This work provides a unified framework for understanding how sensory evidence controls actions through global neural mechanisms, linking concepts from decision-making and motor control fields.