In recent years, there has been a paradigm shift in experimental neuroscience, using emerging technologies to ‘close the loop’ around the nervous system. These experiments measure or stimulate neural activity in the brain of awake, behaving animals based on behavioral or neural variables analyzed in real time. Advancements in position tracking and miniaturized sensors enable neural stimulation to be applied based on complex behavioral or physiological variables. Machine learning can predict and validate optimal behavioral stimuli that elicit a desired neural response, and animals can even be trained to elicit specific neural patterns for reward. Advancements in simultaneous neural recording and stimulation through electrical, optical, acoustic, and chemical channels allow neural activity patterns to dictate neural stimulation. This modifies the nature of neural computation in ways that allow us to dissect and model its components. We survey and present these neural closed-loop manipulations based on their feedback modes and discuss the resultant scientific advancements and remaining challenges.