Thread by @lei_zhang_lz

Our work, "A network supporting social influences in human decision-making", is out in preprint @biorxiv_neursci!
Big thanks to @glascherlab @SysNeuroHamburg
Paper: biorxiv.org/content/10.110…
Suppl.: biorxiv.org/content/10.110…
(1/x)

The question is: in goal-directed learning, how do people adjust their choice and confidence in face of various distributions of others, and how self learning and social learning is integrated to guide the next decisions. (2/x)

To tackle it, we used a probabilistic reversal learning paradigm in a social set up - 5 participants (1 in the MRI scanner) come to the lab to undergo the task simultaneously. Our model reveals dissociated value signals, and captures the behavioral features. (3/x)

We next examine the neural computaions. We find distinct value representation of self learning in vmPFC and social learning in a section of ACC. We also examine the connectivity: regions associated with social info and choice adjustment covary with neural signal of values. (4/x)

We conclude this work by an establishing an interaction of two brain networks related to processing reward and social information, which modulates behavioral adjustment. (5/x)

The modeling work was done with @mcmc_stan, which was always satisfying. As a side, I contributed a bit to the well-received #hBayesDM package, led by @wooyoungahn. It's been updated recently!
the package: github.com/CCS-Lab/hBayes…
(6/x)

Thanks to @glascherlab and @SysNeuroHamburg again, also to people who helped with data collection. Also thank @nathanieldaw for inspring modeling ideas - been a joint work. That's it for now, thanks for reading! (7/7)