Postdoc @UCBerkeley. modeling cognitive processes, improving research practices, & Bayesian everything. she/her
Oakland, CA
Joined November 2018
- Tweets77
- Following166
- Followers227
- Likes190
Beth Baribault retweeted
Hiring full-time predoc RA @MIT @MITSloan! Study decision-making using computational cogsci & behavioral econ! Joint with @amaatouq
apply.interfolio.com/119540
Special encouragement to those underrepresented in computational fields! PREDOC.org
Beth Baribault retweeted
Unsure if science people are still on twitter (?), but if so:
New preprint with @NassarLab ,@Anne_On_Tw, @bbbbaribault, Alex He, Elena Festa and Ed Walsh about age-related changes in reinforcement learning and working memory!
biorxiv.org/content/10.1101/…
Beth Baribault retweeted
New preprint with @nathanieldaw and @DShohamy!
When do people use episodic memory for decisions?
Here we address this question by putting episodic memory in direct competition with learning from trial-and-error. biorxiv.org/content/10.1101/…
so pumped to attend this workshop Saturday @ #RLDM2022!! amazing lineup of computational psychiatry talks 🧠 organized by the fantastic @Rachel_Bedder :)
This Post is from an account that no longer exists. Learn more
using hierarchical Bayesian methods to fit cognitive RL models is more popular than ever!!🔥 Friday @ #RLDM2022 come learn how model development work is **critical** for success in making Bayesian versions of complex cognitive models, like our RL+WM model! @Anne_On_Tw #2.158
bayesian cognitive models are powerful tools to analyze data through the lens of cognitive processes --- but they can be *exceptionally* tricky to use correctly.
in our new preprint, we offer a clear & comprehensive guide to troubleshooting bayesian cognitive models! (1/5)
Check out our new paper from the excellent @bbbbaribault ! An in-depth tutorial on hierarchical bayesian modeling with Stan for cognitive models. @ccnlab
Troubleshooting Bayesian cognitive models: A tutorial with matstanlib psyarxiv.com/rtgew via @OSFramework
we also introduce the matstanlib library (github.com/baribault/matstan…) to help demonstrate good workflow for bayes cog modeling in MATLAB 📉📈📊
matstanlib automates the diagnostic checks & plots required of bayesian models (& was used to generate almost all plots in the paper! 4/5)
how can we model explore-exploit strategies? in my #psynom21 talk (14) TODAY, i'll show how a "modified RL" model offers a satisfying explanation of individual differences in E-E strategies for the observe or bet task! 💻 Motivation, Reward, & Cog Processes session @ 9:00 am CT !
many thanks to my collaborators on this talk, @ManonIronside, Sheri Johnson, and Anne Collins @ccnlab !
not attending #psynom21, or had a session conflict? no problem! i've also uploaded the talk to youtube: youtu.be/7-Wx5qO1NM8 👀
how can we model explore-exploit strategies? in my #psynom21 talk (14) TODAY, i'll show how a "modified RL" model offers a satisfying explanation of individual differences in E-E strategies for the observe or bet task! 💻 Motivation, Reward, & Cog Processes session @ 9:00 am CT !
What do you do when your research software is bugged? Great thread that ends with 4 lessons that are critical for anyone who writes & reuses experiment code, analysis code, …
Also check out this new pub from @ccnlab — using hierarchical Bayesian models (with a cognitive psychometric twist!!) to clarify how reinforcement learning process dynamics develop non-linearly 📈📉 with age!
Check out our new paper here! journals.plos.org/ploscompbi…
with @ccnlab @xia_jimmy @eckstein_maria, @bbbbaribault Sarah Master, Linda Wilbrecht and Ron Dahl!
Is learning from obvious rewards (+1 vs +0 points) different from learning from novel outcomes (arbitrarily designated “goal” vs “non-goal”)? @smickdougle explains how we can learn from goals almost as well as from rewards, with a little help from executive functions 🧠✅
This Post is from an account that no longer exists. Learn more
charming blog post by Clintin talking about 👽 & his excellent new paper on psychological theory development (link in post!)
This Post is from an account that no longer exists. Learn more
