Neural Computations Underlying Causal Structure Learning

Download:

• Paper
• Task code
• Data analysis code

Abstract:

Behavioral evidence suggests that beliefs about causal structure constrain associative learning, determining which stimuli can enter into association, as well as the functional form of that association. Bayesian learning theory provides one mechanism by which structural beliefs can be acquired from experience, but the neural basis of this mechanism is poorly understood. We studied this question with a combination of behavioral, computational, and neuroimaging techniques. Male and female human subjects learned to predict an outcome based on cue and context stimuli while being scanned using fMRI. Using a model-based analysis of the fMRI data, we show that structure learning signals are encoded in posterior parietal cortex, lateral prefrontal cortex, and the frontal pole. These structure learning signals are distinct from associative learning signals. Moreover, representational similarity analysis and information mapping revealed that the multivariate patterns of activity in posterior parietal cortex and anterior insula encode the full posterior distribution over causal structures. Variability in the encoding of the posterior across subjects predicted variability in their subsequent behavioral performance. These results provide evidence for a neural architecture in which structure learning guides the formation of associations.

Figure 5A: Learning of abstract causal structure ($KL_{structures}$) above and beyond learning of specific cause-effect relationships ($KL_{weights}$)

Figure 6A: Posterior beliefs about abstract causal structure after receiving new information

Citation

Tomov, M. S., Dorfman, H. M., Gershman, S. J. (2018). “Neural Computations Underlying Causal Structure Learning.” Journal of Neuroscience, 38(32), 7143-7157. https://doi.org/10.1523/JNEUROSCI.3336-17.2018.

@article{tomov2018neural,
	author = {Tomov, Momchil S. and Dorfman, Hayley M. and Gershman, Samuel J.},
	title = {Neural Computations Underlying Causal Structure Learning},
	volume = {38},
	number = {32},
	pages = {7143--7157},
	year = {2018},
	doi = {10.1523/JNEUROSCI.3336-17.2018},
	publisher = {Society for Neuroscience},
	issn = {0270-6474},
	URL = {https://www.jneurosci.org/content/38/32/7143},
	eprint = {https://www.jneurosci.org/content/38/32/7143.full.pdf},
	journal = {Journal of Neuroscience}
}