Paper: | RS-2B.2 |
Session: | Late Breaking Research 2B |
Location: | Late-Breaking Research Area |
Session Time: | Friday, September 7, 19:30 - 21:30 |
Presentation Time: | Friday, September 7, 19:30 - 21:30 |
Presentation: |
Poster
|
Publication: |
2018 Conference on Cognitive Computational Neuroscience, 5-8 September 2018, Philadelphia, Pennsylvania |
Paper Title: |
Passive forgetting or selective attention? Comparing two models of learning in multidimensional environments |
Authors: |
Guy Davidson, Minerva Schools, United States; Angela Radulescu, Yael Niv, Princeton University, United States |
Abstract: |
Using a multidimensional reinforcement learning task in which one of three dimensions determines reward, previous work showed that cognitive models incorporating passive decay of the values of unchosen options explained subject choice data better than competing models (Niv et al., 2015). More recently, models that assume attention-weighted reinforcement learning were shown to predict the data equally well (Leong et al., 2017). We investigate whether the two models, which suggest different cognitive processes, explain the same aspect of the data, or rather different, complementary aspects. We show that combining the two models improves the overall average fit, suggesting that these two mechanisms explain separate components of the variance. We showcase how each model helps explain different trial dynamics depending on the progress a subject has made in learning the task, going beyond BIC and average cross-validated likelihood and employing a trial-by-trial analysis of choice predictions to uncover subtle differences between models that perform equally well on average. Our results show that attention-weighted learning predicts choice substantially better in trials immediately following the point at which the subject has successfully learned the task, while passive decay better accounts for choices in trials further into the future relative to the point of learning. |