Strategies and representations in physical inference
- Kevin Smith, Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, Massachusetts, United States
- Josh Tenenbaum, Brain and Cognitive Sciences, MIT, Cambridge, Massachusetts, United States
- Erin Anderson, Psychology, Northwestern University, Evanston, Illinois, United States
- Susan Hespos, Psychology, Northwestern University, Evanston, Illinois, United States
- Lance Rips, Psychology Dept., Northwestern University, Evanston, Illinois, United States
- Chaz Firestone, Psychological and Brain Sciences, Johns Hopkins University, Baltimore, Maryland, United States
- Jessica Hamrick, DeepMind, London, United Kingdom
AbstractHuman cognition is not thought to be a monolithic entity, but instead is often considered to include a range of strategies and representations that can be applied selectively in different domains. For instance, we are thought to have both intuitive and deliberative processes for decision making (Kahneman, 2011) and both approximate and precise representations of numeracy (Feigenson, Dehaene, & Spelke, 2004). But how do these different strategies and representations develop, and how do we select which ones to use?