This was an amazingly fun collaboration that began at the AC Summer School @GatsbyUCL, and I can't wait to see where it takes us next! Check out our spotlight paper and poster at NeurIPS. Grainy picture from presenting the first version of this work at @CosyneMeeting attached!

It was a pleasure teaching at the Computational Psychology and AI Summer Schools for undergraduate students. Thanks to @lmhoxford @UniofOxford for hosting! @VedanthNath, it was great having you in the lectures - I look forward to seeing what you are up to in the future!

Remarkably, we find that this individual variation in behavior correlates well with PCs extracted from anxiety & depression and compulsivity transdiagnostic factor scores. We hope these findings can pave the way for using ANNs to study healthy and pathological meta-control! (4/4)

We perturb the hidden representations of the meta-RL networks along the axis used for APE prediction. When perturbed systematically, the models replicate human individual differences in performance across levels of controllability (3/4)

We ask humans and neural networks to complete observe or bet task variants that require adapting to changes in controllability. Meta-RL trained neural networks only match human performance when explicitly trained to predict APEs, mirroring error likelihood prediction in ACC (2/4)

Excited that the preprint for the work from my first two years of PhD at @summerfieldlab is out! In this work, we examine the role of action prediction errors (APEs) in cognitive control: https://t.co/F5tF4NhkvN (1/4)

The final work is ours: deep RL models of meta-control. By learning predictive representations of their own control, these networks distinguish situations requiring different in-context updates. We presented this work at CCN 2023 and will release a longer paper soon - stay tuned!

Needing no introduction, the second work we review is deep meta-RL. Along with large language models, these architectures have the ability to learn solutions that adapt to new situations in-context, effectively "learning-to-learn" (3/4)

A key aspect of cognitive flexibility is the ability to adapt general solutions to specific conditions. By separating learning into habit-driven and goal-driven streams, the first work we review accomplishes this in end-to-end trained nets, capturing many cognitive effects (2/4)

Excited that our new position piece is out! In this article, @summerfieldlab and I review three recent advances in using deep RL to model cognitive flexibility, a hallmark of human cognition: https://t.co/FX52syfUCr (1/4)

Wondering how representational similarities influence our perception of novelty and novelty-driven behaviors? Then don't miss my talk this Friday in #Cosyne2024! Looking forward to great questions and discussions! 🤩😊

✨ As 2023 comes to a close, we reflect back on what an incredible year it’s been for the lab ✨ We had 4 big papers that mark 🔥 milestones for our group: - #CEBRA @cebraAI @Nature - #AmadeusGPT #neurips2023 - modeling proprioception @elife - SOTA pose estimation #ICCV2023 ⬇️

It was a lot of fun presenting my first PhD work at @CogCompNeuro this weekend! Next stop: the Analytical Connectionism Summer School at @GatsbyUCL #AC2023

I had a great time at the @CIFAR_News #DLRL summer school last week! What a fantastic opportunity to learn about cutting-edge methods in ML and explore the fantastic research environment at @Mila_Quebec

@TrackingActions @TrackingPlumes Thanks also to @pranavm42 and the other co-authors who made this work possible - @pranavm42 I'm glad we managed to meet and wrap the project up in person together as work started before and then carried on throughout the entire pandemic!

So many thanks to @TrackingActions and @TrackingPlumes for having hosted me and for their amazing mentorship! This was my first large scientific project, and working with them was a fantastic experience and great introduction to scientific work and practice

So glad that this work is finally out! In https://t.co/hbSXwXtslr, we show how task-driven modelling can be used to investigate the computational role of the proprioceptive system by training neural nets and comparing emergent representations side-by-side for different hypotheses

🥳 @akaijsa & @pranavm42 et al introduce task-driven modelling of the proprioceptive system, now out in @eLife! Our work  combines deep learning & biomechanics to test theories of💪sensory representations https://t.co/3AuKKCvhNx