Collins Lab

📢 New preprint! How do humans learn from arbitrary, abstract goals? We show that, when goal spaces can be compressed, costly working-memory processes give way to internalized reward functions, enabling efficient goal-dependent RL. @Anne_On_Tw @ccnlab arxiv.org/abs/2509.06810

Exciting news - early bird registration is now open for #RLDM2025! Register now: forms.gle/QZS1GkZhYGRFKd… Register now to save €100 on your ticket. Early bird prices are only available until 1st April.

And, a reminder: the deadline to submit an extended abstract to RLDM is this coming Wednesday (Jan. 15)! Details at rldm.org/call-for-abstr…

The RLDM workshop list is now up! Have a look: rldm.org/call-for-works… More workshop details coming soon :) There were so many wonderful submissions---this was really tough for the committee. Huge thanks to all involved, and looking forward to seeing folks in Dublin

*Call for Abstracts* Submit your abstracts on "learning and decision-making over time to achieve a goal" to #RLDM2025 Successful applications will be selected for poster or oral presentation to an interdisciplinary audience. Deadline: Jan 15 Learn more: rldm.org/call-for-abstr…

📷 Calling all experts! 📷 We're seeking tutorial proposals for topics on "learning and decision-making over time to achieve a goal" Help foster cross-disciplinary exchange at RLDM 2025! 📷 Deadline: December 20th 📷 Learn more & apply: rldm.org/call-for-tutor…

Have ideas for an interdisciplinary workshop? Get involved at RLDM to tackle key challenges, open problems, or controversial topics! 📷 Formats: Talks, hackathons, debates—your choice! 📷 All career stages welcome to lead. 📷️ Apply by 10th Dec: rldm.org/call-for-works…

I finally finished writing my fun RL project during sabbatical in France. I first presented on this a year ago, but I'm still puzzled by the results, and what they mean for human reinforcement learning! Feedback welcome! osf.io/preprints/psya…

How do we decide which goals to pursue? Here, we introduce the notion of *latent learning progress* to explain human goal selection and inspire advances in the development of autotelic machines. osf.io/preprints/psya… @Anne_On_Tw @pyoudeyer @cedcolas @ccnlab @FlowersINRIA 🧵

The last paper of my time in @ccnlab is now out, with @Anne_On_Tw, @xia_jimmy and Ti-Fen Pan! journals.plos.org/ploscompbiol/a…

We are hiring a full-time lab manager for @ccnlab to start this summer! Check out our website for information about how to apply: ccn.berkeley.edu/pages/contact-…

New paper with @Anne_On_Tw introducing a modeling framework for noise fluctuations in decision-making. It can be applied to identify lapses of attention, limit the impact of noisy trials on model fit, and avoid excluding some “noisy” subjects. (1/4) authors.elsevier.com/c/1igK253naxrQ1

The notion that goals are central to human cognition is intuitive, yet learning and decision-making researchers have often overlooked the topic. In this @TrendsCognSci article, @Anne_On_Tw and I propose it’s time to start studying goals in their own right authors.elsevier.com/a/1hjzw4sIRvPN…

We have an opening postdoc position with a strong fMRI background at @ccnlab where we study flexible decision-making and learning in humans using behavioral paradigms and computational approaches. Get in touch if this sounds interesting! More info at aprecruit.berkeley.edu/JPF04096

I am looking to hire a postdoc with strong fMRI expertise for an NIH funded project at @ccnlab. More info here: aprecruit.berkeley.edu/JPF04096 Please RT and pass along to potential applicants.

The value we place on an option depends on the context and available alternatives. But why does getting $1 when the possibilities were 0 or 1 feel much better than when the possibilities were 1 or 10? 🧵👇 1/7

If you're doing hierarchical bayesian modeling for cognitive models, @bbbbaribault Beth's new paper is a must read! Highlighting all the most recent best practices, includes a new @matlab toolbox for visualization psycnet.apa.org/record/2023-57…

I am looking to recruit a postdoc for my lab @ccnlab at UC Berkeley ccn.berkeley.edu. This is an NIH funded position, for a methods oriented project - we are looking for applicants with strong quantitative skills. Please forward to potential applicants!

We will be hiring a new lab manager for @ccnlab to start this summer! Interested candidates should email ccnlab@berkeley.edu with their CV and a short cover letter, and we will follow up with additional information when available. More here: ccn.berkeley.edu/pages/contact-…

Our new paper is out! @milenamr7, with @ar__zou and @Anne_On_Tw, explores how choice type impacts human reinforcement learning: doi.org/10.1162/jocn_a…