Are you new to the field of Computational Psychiatry or just looking for resources on applying Bayesian models of cognition to behavioural data? Then check out our new paper "Bayesian Workflow for Generative Modeling in Computational Psychiatry": https://t.co/GQUIIcUdhr 1/6

How do we process uncertainty, and how do neurometabolites shape this?🧠Here we use a combination of computational modeling + 7T MRS to answer these questions: https://t.co/VsfExNtfT3 @PTWaade @chmathys @rikepetzschner @oeparsons @cat_rua @sbaroncohen @beckyneuroet et al 🧵(1/4)

Early bird registrations for the 2025 Brain Connectivity Meeting now open June 20 (arrive) until June 22 on the beautiful North Stradbroke Island (Minjerribah), near Brisbane (location of OHBM) https://t.co/GJbiNkK5oQ

#SAMBAlab Paper alert, written by @D_Atanassova et al. Exploring when to exploit: The cognitive underpinnings of foraging-type decisions in relation to psychopathy. w/@cognemo_andreea,@chmathys & others. @SSSpsychopathy @SSSPStudents @decisioncenter👇🏾👀 https://t.co/n7W6zkwxKB

We are indescribably sad to announce the passing of our dear friend and colleague Kristian Tylén His presence at the IMC will be forever missed and his impact on us all is beyond words Our thoughts are with his wife and children in this difficult time https://t.co/v810Wxqjkz

Super excited to share our new preprint, led by my supervisor Franziska Knolle in collaboration with Verena Demler, @Lucy_MacG & @chmathys: 💡Guided by Expectations: Overweighted Semantic Priors in Schizotypy and their Links to Glutamate https://t.co/ffxyc4B9DP A thread 🧵:

Folks in 🇩🇰 & beyond! On Oct 1st, we are hosting (@interact_minds) a workshop exploring research at the intersection of NLP x social sciences, cognition, & humanities, with 3 awesome keynotes by @dirk_hovy, @maria_antoniak & @sandropezzelle. Register here:

Great to see this out - congratulations @D_Atanassova !

🧵Why do people with psychopathic traits stick to maladaptive patterns of behaviour? The answer: a cognitive rigidity mechanism prompts them to ignore new evidence and maintain prior expectations in pain learning situations. 1/3 https://t.co/GhzrBMeFDq

@PTWaade This has been a long haul – the main work was done as part of @lilwebian's PhD at the #TNU @ETH_en, with Klaas and me. Years later, Nicolas, @PTWaade and Anna came onboard and made shiny toolboxes from our initial codebase. First preprint May 2023. Glad to finally share it! 10/

@PTWaade We also provide some examples for the kinds of models you can now build within the HGF framework – from global volatility beliefs, multimodal inference, multisensory cue combination over time, etc etc – check it out! https://t.co/oxzCHeoW3o 9/

@PTWaade There’s plenty more in the preprint. For example, we find that the message passing implied by value coupling in the HGF is mostly equivalent to what has been proposed for PC, with small, but interesting differences. 8/

Even better, we implemented the gHGF in two(!) openly available toolboxes for you: 1. In Python (led by Nicolas Legrand): https://t.co/Sb98wJPYnE 2. In Julia (led by @PTWaade): https://t.co/wn4WzH7CEM 7/

This modular architecture of the generalised HGF means that it’s now super simple to build your own models under the HGF – you can literally add or remove nodes in your belief network and don’t need to worry about deriving any new equations! 6/

Here, we generalise the HGF to encompass both value and volatility coupling, all under the same framework, with one-step belief updates driven by precision-weighted prediction errors, and a modular belief network architecture. 5/

In contrast, PC typically focuses on hierarchies in which higher level beliefs concern the *value* of lower levels. This is useful for understanding how beliefs about lower-level features depend on context (higher-level beliefs) (e.g., Rao & Ballard, 1999). 4/

In current HGF models, higher level beliefs encode the *volatility* of lower levels. This is motivated by the observation that learning should be influenced by environmental uncertainty: if the world is currently changing faster (volatile), the agent needs to learn faster. 3/

Predictive coding and the HGF are algorithms where messages are passed between the nodes of hierarchical Bayesian models, but they tend to assume different kinds of belief hierarchies: volatility coupling (HGF) versus value coupling (PC). What does that mean?? 2/

Here's a copy of @lilwebian's thread on the generalized HGF https://t.co/oxzCHeoW3o 👇 1/

If you want to have a chat about this, meet me today at #IWAI2024 or the rest of the week at the #CPC in Zurich. There, we will also present the new toolboxes implementing all of this: https://t.co/EBXVW9iIXx (led by Nicolas) and https://t.co/wLA4jPSMxI (led by @PTWaade)