Timezone: »

The rat as particle filter
Nathaniel D Daw · Aaron Courville

Wed Dec 05 05:20 PM -- 05:30 PM (PST) @ None

The core tenet of Bayesian modeling is that subjects represent beliefs as distributions over possible hypotheses. Such models have fruitfully been applied to the study of learning in the context of animal conditioning experiments (and anologously designed human learning tasks), where they explain phenomena such as retrospective revaluation that seem to demonstrate that subjects entertain multiple hypotheses simultaneously. However, a recent quantitative analysis of individual subject records by Gallistel and colleagues cast doubt on a very broad family of conditioning models by showing that all of the key features the models capture about even simple learning curves are artifacts of averaging over subjects. Rather than smooth learning curves (which Bayesian models interpret as revealing the gradual tradeoff from prior to posterior as data accumulate), subjects acquire suddenly, and their predictions continue to fluctuate abruptly. These data demand revisiting the model of the individual versus the ensemble, and also raise the worry that more sophisticated behaviors thought to support Bayesian models might also emerge artifactually from averaging over the simpler behavior of individuals. We suggest that the suddenness of changes in subjects' beliefs (as expressed in conditioned behavior) can be modeled by assuming they are conducting inference using sequential Monte Carlo sampling with a small number of samples --- one, in our simulations. Ensemble behavior resembles exact Bayesian models since, as in particle filters, it averages over many samples. Further, the model is capable of exhibiting sophisticated behaviors like retrospective revaluation at the ensemble level, even given minimally sophisticated individuals that do not track uncertainty from trial to trial. These results point to the need for more sophisticated experimental analysis to test Bayesian models, and refocus theorizing on the individual, while at the same time clarifying why the ensemble may be of interest.

Author Information

Nathaniel D Daw (New York University)

Nathaniel Daw is Assistant Professor of Neural Science and Psychology and Affiliated Assistant Professor of Computer Science at New York University. Prior to this he completed his PhD in Computer Science at Carnegie Mellon University and pursued postdoctoral research at the Gatsby Computational Neuroscience Unit at UCL. His research concerns reinforcement learning and decision making from a computational approach, and particularly the application of computational models to the analysis of behavioral and neural data. He is the recipient of a McKnight Scholar Award, a NARSAD Young Investigator Award, and a Royal Society USA Research Fellowship.

Aaron Courville (U. Montreal)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors