A new research Article in the journal by Sherry Dongqi Bao and colleagues: www.nature.com/articles/s44... (in Press) @sherrydongqi.bsky.social @weijima.bsky.social

New lab preprint! Computational psychiatry assumes anxiety arises from domain-general biases. We argue this is mistargeted: trait anxiety is generated by person-specific world models, which requires a fundamentally different approach to task design. 🧵 https://osf.io/preprints/psyarxiv/98ktg_v3

What an awesome #cpconf @ Yale! I am SO proud of my amazing students 👏👏 @noa-yadid.bsky.social won a poster award 🎉 on ITIs impact on learning @adipizanti.bsky.social lectured on our DYNA-checking model of worry @dyavetz.bsky.social presented a model of planning capacity & efficiency in aging

Very happy to be in New Haven at #cpconf with a bunch of the Translational Lab presenting! 🧵

Very grateful to be able to travel to an academic conference in the US by train! A smooth 5-hr "high-speed" ride from DC to New Haven non-stop, lots of work done, and very excited to catch up with friends and colleagues at #cpconf #cpconf2026 🚨 My lab will be presenting 4 posters, details below 🔽

We're excited to share the lab's findings at #Cpconf See 🧵 below & come by! C71 + poster spotlight! @adipizanti.bsky.social finds overthinking (using MB planning more effortfully without any benefit) is related to worry in 2 large datasets, and explains it with a new DYNA + internal checking model:

Excited to be heading to #cpconf in New Haven and looking forward to what looks like a really brilliant programme! We have some exciting new work being presented by our brilliant students and postdocs 👇👇👇

How does the human brain extract abstract structure from experience? We show that sensory-independent relational representations emerge in mPFC over several days. Work with @mgarvert.bsky.social and @behrenstimb.bsky.social now out in @currentbiology.bsky.social: www.cell.com/current-biol...

Most of the statistical tests you learned in stats class — t-test, ANOVA, correlation — are actually special cases of a single thing: linear regression. Ch 7 of Experimentology argues that thinking in models, not tests, is more flexible and a better foundation for theory. 🧵 experimentology.io

The riddle riddle Human common sense prevails

Bella Fascendini@bellafascendini.bsky.social · last mo.

New paper! w/ @cocoscilab.bsky.social🧵Can large language models reason flexibly, or have they learned what reasoning looks like? We introduce a new paradigm to test this question—the riddle riddle—and find that humans and LLMs show opposite patterns of performance. 📜

Studies on risk aversion show that behavior shifts depending on the domain (sometimes a lot). Yet, when modeling risk aversion, we remove prior knowledge and fit a single parameter. Why not model directly how our prior knowledge (world model) shapes task representation & task-specific risk aversion?

If anyone knows of a “risky test” the popular HiTOP produces, please point me to it. I mean an out of sample unintuitive prediction that reflects the strength of the theory.

—— Open Mind is MIT Press’s Diamond Open Access Cognitive Science journal. We now have our first Impact Factor: 2.9 as of June 2026 Submit your papers to Open Mind direct.mit.edu/opmi

🚨Preprint alert🚨 Research on discourse incoherence (e.g., in psychosis) usually focuses on patients' language, but what happens to people who interact with misaligned partners? Using a novel approach, we show evidence for reciprocity in semantic misalignment. 🧵👇 osf.io/preprints/ps...

Computational psychiatry needs more developmentalists. If I were to bet, I'd say, there's a critical window of time where core beliefs (hyperpriors) form through learning early in life. These shape our world models that allow us to make sense of new data through the causal knowledge we've accrued.

Theories let us make bets. Predicting "happiness goes up with money" is a roulette bet on red — half of outcomes confirm it. Predicting a specific log-linear relationship is more like betting on a single number. "Risky tests" (specific bets) are how we actually adjudicate between theories.

🎉New positions!🎉 Postdoc and Research technician openings in our lab @yale for computational psychiatry research. Technician position a great stepping stone to grad school. More info: rutledgelab.org/positions Work w/big data from our smartphone apps and large clinical samples happinessquest.app

Postdoc and Research Technician Positions — Rutledge Lab

Now hiring a postdoc and a research technician with an interest in neuroeconomics or computational psychiatry.

rutledgelab.org

out now in Psych Review, a project many years in the making: "Reverse Engineering the Centered Self" bit.ly/4wp9fgN

How can 🍪 help us understand anhedonia? What started as a small project is now (> 2 years after the preprint, with revision deadlines landing exactly during my move from Tübingen to Bonn and even during a monastery-based PhD retreat) finally out in Cell Rep Med 🔽🧵 www.cell.com/cell-reports...

Blunted anticipation, but not consummation, of food rewards in depression

Schulz et al. show that anhedonia in depression, as measured by a scale often equated with loss of pleasure, reflects reduced anticipation of food rewards rather than impaired enjoyment itself. Linkin...

cell.com

Please give us cumulative effect size over time plots, hopefully also explaining how the model is improved gradually 🤞 I just gave a philosophy of science talk to our PhD students, asking for this from the field. Great to see part of this center stage in an intro text!

Mike Frank@mcxfrank.bsky.social · 3mo ago

Many of us were taught experiments are for testing hypotheses. In Ch 1 of Experimentology, our free, open methods textbook, my coauthors and I argue differently: experiments are for estimating the magnitude of causal effects. This reframing has important consequences. 🧵 experimentology.io

It's a gift to do what you love, and it pays off.

Natural History Museum, London@nhm-london.bsky.social · 3mo ago

Join us in wishing a massive happy 100th birthday to the man who has inspired millions of people around the world to care about nature and look after our planet - the incredible Sir David Attenborough! 🎉

New preprint! Last year we showed humans adapt attention in a dimension-specific way when multiple things compete for attention. Now we read out the neural dynamics of this adaptation. doi.org/10.1101/2025.11.16.688701 w/ @mikefreund.bsky.social @theazalabak.bsky.social @wouterkool.bsky.social

This research “gives us a chance to look into each other’s universes and ask the question of why we process the world the way we do,” says Sharp. “To me, this is the origin of our humanity.” A little article in Aperio magazine on our lab's work. azrielifoundation.org/aperio-magaz...

Cracking the Code of the Anxious Mind - The Azrieli Foundation

Paul Sharp treats the brain like a computer to understand why we get stuck in doom loops — and how we can reprogram our mental software

azrielifoundation.org

🚨 Our lab's first paper is published! The claim: anxiety distorts not just what looks threatening, but the task models people use to plan. 🧵 This tiny study provides preliminary support that world model variation is critical to explaining anxiety.

🚨 New preprint lead by the unstoppable @luiantaverra.bsky.social 👇 Does avoidance distort our beliefs you ask? Do anxious traits modulate this relations? Find out in the paper ^^. Also check out the new gamified task that Luianta developed: luiantav.github.io/dog-island/

Luianta Verra@luiantaverra.bsky.social · 3mo ago

New preprint! How do threat beliefs drive avoidance & what happens when avoidance prevents us from verifying them? New work showing a bidirectional link between beliefs and avoidance! Lucky to work w @ondrejzika.bsky.social @nicoschuck.bsky.social @tobywise.bsky.social osf.io/preprints/ps...

🚨 Our lab's first paper is published! The claim: anxiety distorts not just what looks threatening, but the task models people use to plan. 🧵 This tiny study provides preliminary support that world model variation is critical to explaining anxiety. onlinelibrary.wiley.com/doi/10.1111/...

Anxiety is Associated With Biases in Task Generalization

Anxiety is known to distort threat perception, but can it also distort the task representations we use to plan? This study shows that individuals high in trait worry systematically undergeneralize ac...

onlinelibrary.wiley.com

Top of my reading list 💫

Marcelo Mattar@marcelomattar.bsky.social · 4mo ago

New Annual Review with @nathanieldaw.bsky.social: “Planning in the Brain: It's Not What You Think It Is.” We argue that the brain's 'planning' machinery is mostly used for learning from simulated experience, and that thinking prospectively at decision time is just one special case of this process.

I don't think risk aversion in trait anxiety is conceptualized well in most comp models. My bet is risk aversion is in part domain-specific. We need to first infer world models - hierarchical prior knowledge schema - to really measure risk aversion we care about. See: www.cell.com/trends/cogni...

Anxiety involves altered planning

Clinicians have suggested but not shown how anxiety involves altered planning. Here, I synthesize and extend computational models of planning in a framework that can be used to explain planning biases...

cell.com

LLM agents are a serious problem for online experiments. It is very easy to use them and very hard to spot them. What can researchers do? With @brendenlake.bsky.social, we suggest detecting LLMs based on their lack of human cognitive constraints in our #CogSci2026 paper: arxiv.org/abs/2604.00016